{
 "_meta": {
  "dataset": "Cyclotron Info — builder FAQ",
  "url": "https://cyclotroninfo.com/learn/faq/",
  "license": "CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)",
  "attribution": "Cyclotron Info (cyclotroninfo.com)",
  "exported": "2026-11-22",
  "note": "Plain-language answers for amateur cyclotron builders, each citing the Design Guide rules and documented builds behind it; the short answer assumes only the stated baseline (https://cyclotroninfo.com/learn/faq/#baseline). Not a substitute for the cited sources or for the safety page.",
  "count": 52
 },
 "entries": [
  {
   "id": "goal-can-and-should",
   "question": "Can an amateur actually build a cyclotron? Should I?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 1,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Yes, and the record says who did it and how long it took. The build census lists 27 amateur and school cyclotron projects since 1947. 8 of them demonstrated or routinely operated a beam; the rest stalled, are still in progress, or simply left no further record, which is not the same as failing. Most of the finished machines had an institution behind them, a college or a school. A few were built at home, and one of those logged first beam within about a year of its start (census). Should you? Probably not, for reasons that have nothing to do with physics. It is expensive. The work is mostly tedious and spans half a dozen trades, vacuum, RF and machining among them, across years with little visible progress. The result is a needle on a meter, and there is little to do with it afterward (should-you-build-one page). Before buying iron, decide the purpose, concretely enough to fail (see What should count as a first success? Do I need to get the beam out of the chamber, or can I demonstrate acceleration with a detector inside? ), and whether you are building alone or with an institution behind you. Then pick a first milestone that pays off in months.",
   "short_answer_markdown": "Yes, and the record says who did it and how long it took. The build census lists 27 amateur and school cyclotron projects since 1947. 8 of them demonstrated or routinely operated a beam; the rest stalled, are still in progress, or simply left no further record, which is not the same as failing. Most of the finished machines had an institution behind them, a college or a school. A few were built at home, and one of those logged first beam within about a year of its start (census). Should you? Probably not, for reasons that have nothing to do with physics. It is expensive. The work is mostly tedious and spans half a dozen trades, vacuum, RF and machining among them, across years with little visible progress. The result is a needle on a meter, and there is little to do with it afterward (should-you-build-one page). Before buying iron, decide the purpose, concretely enough to fail (see goal-first-success), and whether you are building alone or with an institution behind you. Then pick a first milestone that pays off in months.",
   "sources": [],
   "links": [
    "https://cyclotroninfo.com/learn/should-you-build-one/",
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#skills",
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#timeline",
    "https://cyclotroninfo.com/builds/",
    "https://cyclotroninfo.com/learn/choosing-your-machine/#brief"
   ],
   "refs": [
    "goal-first-success",
    "lit-what-to-read",
    "safety-before-testing"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#goal-can-and-should"
  },
  {
   "id": "goal-first-success",
   "question": "What should count as a first success? Do I need to get the beam out of the chamber, or can I demonstrate acceleration with a detector inside?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 2,
   "tags": [
    "plan",
    "prove"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "A detector inside the chamber is the standard of success. None of the amateur or school machines in the census is documented to have run an external beam. The first success is a resonance peak on an internal collector that behaves like a beam. Sweep the magnetic field at a fixed RF frequency, and the collector current rises at the field where the ions cross the gap in step with the RF every turn. A stray current that reaches the plate by some other route does not depend on that match, and Wouters’ 1949 checks use the difference. The evidence becomes strong when the peak is sharp, responds to hydrogen pressure, and dies when field or frequency is moved a little off the match. Write two things into the plan as the definition of done: that peak passing those checks, and the species, which kind of ion it was. Name the species as candidates, or identify it as far as the evidence allows. Energy can stay computed, labeled as computed. Extraction belongs in a later plan, if ever, and only after the internal beam is understood.",
   "short_answer_markdown": "A detector inside the chamber is the standard of success. None of the amateur or school machines in the census is documented to have run an external beam. The first success is a resonance peak on an internal collector that behaves like a beam. Sweep the magnetic field at a fixed RF frequency, and the collector current rises at the field where the ions cross the gap in step with the RF every turn. A stray current that reaches the plate by some other route does not depend on that match, and Wouters' 1949 checks use the difference. The evidence becomes strong when the peak is sharp, responds to hydrogen pressure, and dies when field or frequency is moved a little off the match. Write two things into the plan as the definition of done: that peak passing those checks, and the species, which kind of ion it was. Name the species as candidates, or identify it as far as the evidence allows. Energy can stay computed, labeled as computed. Extraction belongs in a later plan, if ever, and only after the internal beam is understood.",
   "sources": [
    "dg-1431",
    "dg-506",
    "dg-881"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-extraction/#the-default",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/beam-measurement/#energy",
    "https://cyclotroninfo.com/learn/choosing-your-machine/#purpose",
    "https://cyclotroninfo.com/legal/"
   ],
   "refs": [
    "beam-documenting-the-claim",
    "beam-is-it-working",
    "beam-which-ion",
    "lit-which-rules-apply"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#goal-first-success"
  },
  {
   "id": "goal-how-it-fits-together",
   "question": "How do the particle, the energy, the magnet size, the field and the RF frequency fit together? Which of them can I actually choose?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 3,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Start from the purpose, which sets the energy you need, tens to hundreds of keV for the machines this FAQ is about; then you choose about two more things, and physics constrains the rest. Choose the particle. Choose the magnetic field, or the pole size, the diameter of the iron faces the ions circle between, because with the energy fixed each sets the other. The highest energy is set by the field and the radius the last turn can use. Pick any two of energy, field and radius and the third follows (cyclotron-energy calculator). The RF frequency then follows from the field and the particle, to a first approximation with no freedom at all. An ion circles at a rate set by the field and its charge and mass, 15.2 MHz per tesla for protons, and the RF has to run at that rate. The dee voltage, the RF voltage across the gap, is a separate choice. Each crossing adds an energy set by that voltage, so it largely determines how many turns the ion takes to reach the final energy. Every turn is more distance travelled through the residual gas, the gas left in the chamber, so the turn count is what the vacuum requirement is estimated from (path-length calculator).",
   "short_answer_markdown": "Start from the purpose, which sets the energy you need, tens to hundreds of keV for the machines this FAQ is about; then you choose about two more things, and physics constrains the rest. Choose the particle. Choose the magnetic field, or the pole size, the diameter of the iron faces the ions circle between, because with the energy fixed each sets the other. The highest energy is set by the field and the radius the last turn can use. Pick any two of energy, field and radius and the third follows (cyclotron-energy calculator). The RF frequency then follows from the field and the particle, to a first approximation with no freedom at all. An ion circles at a rate set by the field and its charge and mass, 15.2 MHz per tesla for protons, and the RF has to run at that rate. The dee voltage, the RF voltage across the gap, is a separate choice. Each crossing adds an energy set by that voltage, so it largely determines how many turns the ion takes to reach the final energy. Every turn is more distance travelled through the residual gas, the gas left in the chamber, so the turn count is what the vacuum requirement is estimated from (path-length calculator).",
   "sources": [
    "dg-115",
    "dg-1375",
    "dg-1386",
    "dg-1387",
    "dg-1398",
    "dg-269"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#resonance-condition",
    "https://cyclotroninfo.com/learn/choosing-your-machine/#design-point",
    "https://cyclotroninfo.com/learn/magnet-design/#field-spec",
    "https://cyclotroninfo.com/calculators/cyclotron-energy/",
    "https://cyclotroninfo.com/calculators/path-length/",
    "https://cyclotroninfo.com/learn/exercises/find-the-resonance/",
    "https://cyclotroninfo.com/learn/exercises/more-voltage-same-energy/",
    "https://cyclotroninfo.com/learn/exercises/what-the-gas-costs/"
   ],
   "refs": [
    "rf-why-resonate",
    "source-where-does-it-go",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#goal-how-it-fits-together"
  },
  {
   "id": "goal-electrons-or-other-ions",
   "question": "Could I accelerate electrons or other ions instead of protons? Would any of those be easier for an amateur?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 4,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "A cyclotron will accelerate anything charged, and changing the particle does not make the machine much easier. The frequency, the energy and the source change. The magnetic force that turns an ion grows with its charge, and the turning it produces shrinks with its mass. At the same magnetic field the circling rate, and so the RF frequency, scales with charge over mass. The same force fixes the momentum an ion carries at a given radius, so momentum grows with charge, and energy is momentum squared over twice the mass. Energy at the same field and radius therefore scales with charge squared over mass ( Humphries (1986), p. 524 ; see the table). The lower frequency of a heavier ion is a different RF design, not automatically an easier one. For a first machine: protons or H₂⁺, a hydrogen molecule with one electron removed, from hydrogen, because the sources, the RF frequency range and the published record all assume them. Expect a species mix, more than one kind of ion, and plan to identify it. Helium if hydrogen handling is the obstacle. Deuterium only inside a radiation plan. Electrons only if the goal is to watch a few-keV orbit rather than to build an accelerator.",
   "short_answer_markdown": "A cyclotron will accelerate anything charged, and changing the particle does not make the machine much easier. The frequency, the energy and the source change. The magnetic force that turns an ion grows with its charge, and the turning it produces shrinks with its mass. At the same magnetic field the circling rate, and so the RF frequency, scales with charge over mass. The same force fixes the momentum an ion carries at a given radius, so momentum grows with charge, and energy is momentum squared over twice the mass. Energy at the same field and radius therefore scales with charge squared over mass (dg-269; see the table). The lower frequency of a heavier ion is a different RF design, not automatically an easier one. For a first machine: protons or H₂⁺, a hydrogen molecule with one electron removed, from hydrogen, because the sources, the RF frequency range and the published record all assume them. Expect a species mix, more than one kind of ion, and plan to identify it. Helium if hydrogen handling is the obstacle. Deuterium only inside a radiation plan. Electrons only if the goal is to watch a few-keV orbit rather than to build an accelerator.",
   "sources": [
    "dg-1494",
    "dg-1754",
    "dg-269",
    "dg-818"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#relativistic-limit",
    "https://cyclotroninfo.com/learn/ion-sources/#h-minus-h-plus",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/safety/#hazard-activation",
    "https://cyclotroninfo.com/learn/exercises/weigh-the-ion/",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "beam-which-ion",
    "safety-before-beam"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#goal-electrons-or-other-ions"
  },
  {
   "id": "lit-which-rules-apply",
   "question": "How do I know which requirements in a book or paper apply to my cyclotron? One says the magnetic field should be uniform, another says it should decrease with radius, and another says it should increase. Who is right?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 5,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "All three can be right, because they describe different requirements; reading the literature means working out which requirement a sentence serves before deciding whether it binds you. A uniform magnetic field is the ideal for keeping the circling rate constant, since that rate depends on the field. A field weaker at the outer turns than at the centre, decreasing slightly with radius, pushes an ion that strays up or down back toward the middle of the gap. That is vertical focusing, and its cost is a small timing drift called phase slip. In the weaker outer field the ion circles a little more slowly, reaches the gap a little later each turn, and the lag adds up. A field that increases with radius belongs to machines whose revolution time stays the same at every energy even as the ion grows heavier at tens of MeV: isochronous cyclotrons. A first amateur machine is the classical kind. It accepts the few-percent radial decrease because the focusing matters more than the slip at sub-MeV energies, while holding the field nearly constant around each circle. Before adopting or skipping a requirement, ask what machine class, energy, current and purpose it was written for.",
   "short_answer_markdown": "All three can be right, because they describe different requirements; reading the literature means working out which requirement a sentence serves before deciding whether it binds you. A uniform magnetic field is the ideal for keeping the circling rate constant, since that rate depends on the field. A field weaker at the outer turns than at the centre, decreasing slightly with radius, pushes an ion that strays up or down back toward the middle of the gap. That is vertical focusing, and its cost is a small timing drift called phase slip. In the weaker outer field the ion circles a little more slowly, reaches the gap a little later each turn, and the lag adds up. A field that increases with radius belongs to machines whose revolution time stays the same at every energy even as the ion grows heavier at tens of MeV: isochronous cyclotrons. A first amateur machine is the classical kind. It accepts the few-percent radial decrease because the focusing matters more than the slip at sub-MeV energies, while holding the field nearly constant around each circle. Before adopting or skipping a requirement, ask what machine class, energy, current and purpose it was written for.",
   "sources": [
    "dg-003",
    "dg-1378",
    "dg-1727",
    "dg-498",
    "dg-979"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#weak-focusing",
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#relativistic-limit",
    "https://cyclotroninfo.com/learn/beam-dynamics/#tune-diagram",
    "https://cyclotroninfo.com/learn/magnet-design/#field-spec",
    "https://cyclotroninfo.com/design-guide/level/1/",
    "https://cyclotroninfo.com/learn/modeling-tools/#beam-dynamics"
   ],
   "refs": [
    "goal-first-success",
    "lit-what-to-read",
    "magnet-how-good"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#lit-which-rules-apply"
  },
  {
   "id": "goal-copy-and-swap",
   "question": "Can I copy a successful amateur cyclotron but change it to suit the parts I already have? Can I combine parts of different designs, or just scale one down?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 6,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Copy a proven machine if you can. It is the cheapest design decision available, and professional laboratories did it too. Swap parts with care, because a cyclotron’s subsystems are tied together by a few hard relations, and a change in one usually forces changes in two others. The magnetic field sets the RF frequency and, with the radius, the energy. The dee and the chamber around it form a capacitor, and the RF circuit is built around its value. The dee voltage sets how many turns the ions make, and more turns means more path through the leftover gas, so the turn count sets the pressure estimate. Scaling down is possible, but it is not the same machine made smaller. The magnetic field distribution, the pattern of field across the gap, scales cleanly; the RF and the energy requirements change with the geometry and have to be redesigned. When one experienced group moved its working design to a larger magnet, most of the beam did not come with it, an illustration of recommissioning, not a controlled experiment. Whatever you change, redo the design point, the pair of field and radius, from the new part’s measured properties.",
   "short_answer_markdown": "Copy a proven machine if you can. It is the cheapest design decision available, and professional laboratories did it too. Swap parts with care, because a cyclotron's subsystems are tied together by a few hard relations, and a change in one usually forces changes in two others. The magnetic field sets the RF frequency and, with the radius, the energy. The dee and the chamber around it form a capacitor, and the RF circuit is built around its value. The dee voltage sets how many turns the ions make, and more turns means more path through the leftover gas, so the turn count sets the pressure estimate. Scaling down is possible, but it is not the same machine made smaller. The magnetic field distribution, the pattern of field across the gap, scales cleanly; the RF and the energy requirements change with the geometry and have to be redesigned. When one experienced group moved its working design to a larger magnet, most of the beam did not come with it, an illustration of recommissioning, not a controlled experiment. Whatever you change, redo the design point, the pair of field and radius, from the new part's measured properties.",
   "sources": [
    "dg-023",
    "dg-1125",
    "dg-1311",
    "dg-1327",
    "dg-1387",
    "dg-1398",
    "dg-1434",
    "dg-1462",
    "dg-1728",
    "dg-1754",
    "dg-960"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/choosing-your-machine/#copy",
    "https://cyclotroninfo.com/learn/magnet-design/#field-spec",
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "goal-do-i-need-to-simulate",
    "goal-how-it-fits-together",
    "magnet-surplus-mri-nmr"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#goal-copy-and-swap"
  },
  {
   "id": "goal-do-i-need-to-simulate",
   "question": "Do I need to simulate the cyclotron before building it? If a simulation shows one proton spiraling outward, how much does that prove?",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 7,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "No, not in the sense of mastering a beam-dynamics package, a program that follows many ions through the machine. Most documented amateur builds do not report running one. The ones that finished mostly describe closed-form arithmetic, formulas worked by hand or on the site’s calculators, plus one free magnet solver (modeling-tools page). A magnet solver is a program that computes the field a given arrangement of iron and coil produces. Modeling the magnet pays before iron is cut. Copper and steel are the expensive mistake, and the field’s shape across the gap depends on the whole iron geometry, so it cannot be discovered by trial winding. The spiraling proton shows that, within that model’s assumptions and numerical setup, the field and frequency you typed in are consistent, which the resonance relation already told you. It says nothing about whether a real population of ions, born at every point in the RF cycle and at scattered positions, survives to the collector. For a first machine, copy a proven design and model only what you change. Measure what can be measured: the capacitance between the dee and the chamber, and the field at the centre against coil current.",
   "short_answer_markdown": "No, not in the sense of mastering a beam-dynamics package, a program that follows many ions through the machine. Most documented amateur builds do not report running one. The ones that finished mostly describe closed-form arithmetic, formulas worked by hand or on the site's calculators, plus one free magnet solver (modeling-tools page). A magnet solver is a program that computes the field a given arrangement of iron and coil produces. Modeling the magnet pays before iron is cut. Copper and steel are the expensive mistake, and the field's shape across the gap depends on the whole iron geometry, so it cannot be discovered by trial winding. The spiraling proton shows that, within that model's assumptions and numerical setup, the field and frequency you typed in are consistent, which the resonance relation already told you. It says nothing about whether a real population of ions, born at every point in the RF cycle and at scattered positions, survives to the collector. For a first machine, copy a proven design and model only what you change. Measure what can be measured: the capacitance between the dee and the chamber, and the field at the centre against coil current.",
   "sources": [
    "dg-1024",
    "dg-1324",
    "dg-1387",
    "dg-1398",
    "dg-1462",
    "dg-1601",
    "dg-1786",
    "dg-1842"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/modeling-tools/#magnetic-field",
    "https://cyclotroninfo.com/learn/modeling-tools/#beam-dynamics",
    "https://cyclotroninfo.com/learn/magnet-design/#model-first",
    "https://cyclotroninfo.com/learn/exercises/out-of-step/",
    "https://cyclotroninfo.com/learn/exercises/where-ions-are-born/",
    "https://cyclotroninfo.com/calculators/"
   ],
   "refs": [
    "goal-copy-and-swap",
    "magnet-how-good",
    "source-timing-with-rf"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#goal-do-i-need-to-simulate"
  },
  {
   "id": "lit-what-to-read",
   "question": "What should I read? The current literature seems hopelessly complicated and theoretical, and the old literature must be way out of date.",
   "group": "A",
   "group_title": "Before You Build",
   "part": 1,
   "order": 8,
   "tags": [
    "plan"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Start with the construction reports of the 1940s to 1970s, and leave the modern professional literature for later. The reports were written by people building machines close to your size who had to explain to a machinist what to cut next, and what has aged in them is the equipment and the vocabulary. The physics, the design decisions and the failure modes read as current. Most modern professional papers are about machines unlike yours: isochronous (the ions take the same time per turn at every energy), often superconducting (magnet coils cooled until they carry current without resistance), delivering milliamperes of beam. Their open questions are ones a first-beam builder can defer; the documented amateur first beams were built from the old reports, the amateur record and the builders’ own notes. The exception is the modern amateur and teaching record (Rutgers, Houghton, COLUMBUS), which is small and hands-on. Read the classics for the engineering, the amateur record for what it is like at your scale, and this site’s safety and legal pages for the part the classics get wrong, their era’s safety practice.",
   "short_answer_markdown": "Start with the construction reports of the 1940s to 1970s, and leave the modern professional literature for later. The reports were written by people building machines close to your size who had to explain to a machinist what to cut next, and what has aged in them is the equipment and the vocabulary. The physics, the design decisions and the failure modes read as current. Most modern professional papers are about machines unlike yours: isochronous (the ions take the same time per turn at every energy), often superconducting (magnet coils cooled until they carry current without resistance), delivering milliamperes of beam. Their open questions are ones a first-beam builder can defer; the documented amateur first beams were built from the old reports, the amateur record and the builders' own notes. The exception is the modern amateur and teaching record (Rutgers, Houghton, COLUMBUS), which is small and hands-on. Read the classics for the engineering, the amateur record for what it is like at your scale, and this site's safety and legal pages for the part the classics get wrong, their era's safety practice.",
   "sources": [
    "dg-018",
    "dg-1054",
    "dg-1848",
    "dg-1855",
    "dg-2061",
    "dg-254",
    "dg-277",
    "dg-318",
    "dg-387",
    "dg-388",
    "dg-473",
    "dg-960"
   ],
   "links": [
    "https://cyclotroninfo.com/library/",
    "https://cyclotroninfo.com/library/classics/",
    "https://cyclotroninfo.com/learn/study-path/#reading-sequence",
    "https://cyclotroninfo.com/learn/study-path/#working-machines",
    "https://cyclotroninfo.com/library/rutgers-umd/",
    "https://cyclotroninfo.com/library/amateur/",
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#reading-order",
    "https://cyclotroninfo.com/learn/safety/"
   ],
   "refs": [
    "lit-which-rules-apply"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-build/#lit-what-to-read"
  },
  {
   "id": "safety-before-testing",
   "question": "What safety checks, expert review and local permissions do I need before I start testing the subsystems?",
   "group": "B",
   "group_title": "Before You Switch Anything On",
   "part": 2,
   "order": 1,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": true,
   "short_answer": "It depends on which subsystem, and the first job is to say which hazards that particular test brings into the room. So, for the subsystem about to be energized, write down what energy it stores or releases, what that can do to the equipment, what it can do to a person, and what review and protection that calls for. Establish those protections, verify them, complete the review, and only then run the planned test. Two matters cut across everything and are arranged before the first test of any kind: the machine’s mains supply, and the jurisdiction. The mains supply is the building wiring that feeds the machine; it falls under the local electrical code and an electrician’s eye. The jurisdiction is the rules of the place where the machine sits, the radiation-control program among them, and programs differ (legal page). Radiation protection joins the list before any subsystem capable of generating hazardous X-rays is operated, which means sufficiently high-voltage electrodes in vacuum, the dee among them, and not the pump. A high enough voltage across a vacuum gap pulls electrons out of the metal, the gap accelerates them, and electrons stopped in metal give off X-rays. The pump has no such voltage. The reviews that need eyes other than yours are the radiation plan and its survey, the mains side, and the rigging of the magnet. None of them waits for a beam. Wouters’ 1949 instruction frames it: “proper precautions must be taken, even during preliminary testing” ( Wouters (1949), p. 9 ).",
   "short_answer_markdown": "It depends on which subsystem, and the first job is to say which hazards that particular test brings into the room. So, for the subsystem about to be energized, write down what energy it stores or releases, what that can do to the equipment, what it can do to a person, and what review and protection that calls for. Establish those protections, verify them, complete the review, and only then run the planned test. Two matters cut across everything and are arranged before the first test of any kind: the machine's mains supply, and the jurisdiction. The mains supply is the building wiring that feeds the machine; it falls under the local electrical code and an electrician's eye. The jurisdiction is the rules of the place where the machine sits, the radiation-control program among them, and programs differ (legal page). Radiation protection joins the list before any subsystem capable of generating hazardous X-rays is operated, which means sufficiently high-voltage electrodes in vacuum, the dee among them, and not the pump. A high enough voltage across a vacuum gap pulls electrons out of the metal, the gap accelerates them, and electrons stopped in metal give off X-rays. The pump has no such voltage. The reviews that need eyes other than yours are the radiation plan and its survey, the mains side, and the rigging of the magnet. None of them waits for a beam. Wouters' 1949 instruction frames it: \"proper precautions must be taken, even during preliminary testing\" (dg-282).",
   "sources": [
    "dg-1069",
    "dg-1072",
    "dg-1073",
    "dg-1081",
    "dg-1082",
    "dg-1084",
    "dg-1935",
    "dg-282"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/safety/",
    "https://cyclotroninfo.com/learn/safety/#three-rules",
    "https://cyclotroninfo.com/legal/#what-builders-do",
    "https://cyclotroninfo.com/legal/states/",
    "https://cyclotroninfo.com/calculators/lid-deflection/",
    "https://cyclotroninfo.com/design-guide/safety/"
   ],
   "refs": [
    "ground-what-floats",
    "magnet-switch-off-voltage",
    "safety-before-beam",
    "vac-what-pump"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-switch-anything-on/#safety-before-testing"
  },
  {
   "id": "safety-before-beam",
   "question": "Can the machine make radiation, or hurt me some other way, before I've even detected a beam?",
   "group": "B",
   "group_title": "Before You Switch Anything On",
   "part": 2,
   "order": 2,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": true,
   "short_answer": "Yes. The hazards most likely to hurt a builder do not need a beam. A vacuum gap holding tens of kilovolts can emit electrons from its own surfaces and turn them into X-rays with no ion source, no gas and no magnet involved, and the dee gap is such a gap. The power supplies, the stored energy in capacitors and cables, the magnet and the vacuum vessel are hazardous from the day they are powered. “No beam yet” is first a statement about a measurement: it can mean too few ions, no acceleration, beam lost before the collector, poor collection, or an instrument problem. It is not evidence about radiation, and the safety case for a machine rests on a hazard assessment, shielding, access control, instruments whose energy response is known, and the local registration rules, not on whether the electrometer has shown a peak. The electrometer is the meter on the collector. What to do before first power-up: a hazard assessment by subsystem, not by beam status, and it decides the rest. Subsystem means each part of the machine taken on its own. The controls and instruments follow the hazards it identifies. A roughing pump is the first-stage pump that takes the chamber from air pressure down to a rough vacuum. A roughing-pump test and a kilovolt dee in vacuum do not need the same protection plan, and neither gets a free pass ( What safety checks, expert review and local permissions do I need before I start testing the subsystems? ).",
   "short_answer_markdown": "Yes. The hazards most likely to hurt a builder do not need a beam. A vacuum gap holding tens of kilovolts can emit electrons from its own surfaces and turn them into X-rays with no ion source, no gas and no magnet involved, and the dee gap is such a gap. The power supplies, the stored energy in capacitors and cables, the magnet and the vacuum vessel are hazardous from the day they are powered. \"No beam yet\" is first a statement about a measurement: it can mean too few ions, no acceleration, beam lost before the collector, poor collection, or an instrument problem. It is not evidence about radiation, and the safety case for a machine rests on a hazard assessment, shielding, access control, instruments whose energy response is known, and the local registration rules, not on whether the electrometer has shown a peak. The electrometer is the meter on the collector. What to do before first power-up: a hazard assessment by subsystem, not by beam status, and it decides the rest. Subsystem means each part of the machine taken on its own. The controls and instruments follow the hazards it identifies. A roughing pump is the first-stage pump that takes the chamber from air pressure down to a rough vacuum. A roughing-pump test and a kilovolt dee in vacuum do not need the same protection plan, and neither gets a free pass (safety-before-testing).",
   "sources": [
    "dg-1389",
    "dg-1670",
    "dg-1932",
    "dg-1935",
    "dg-1960",
    "dg-2053",
    "dg-559",
    "dg-614"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/safety/#hazard-xrays",
    "https://cyclotroninfo.com/learn/safety/#hazard-electrical",
    "https://cyclotroninfo.com/learn/safety/#three-rules",
    "https://cyclotroninfo.com/learn/shielding/#what-you-emit",
    "https://cyclotroninfo.com/legal/",
    "https://cyclotroninfo.com/design-guide/safety/"
   ],
   "refs": [
    "magnet-switch-off-voltage",
    "rf-sub-kilovolt-dees",
    "rf-watts-to-volts",
    "safety-before-testing"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-switch-anything-on/#safety-before-beam"
  },
  {
   "id": "magnet-switch-off-voltage",
   "question": "I've heard that switching off the magnet coils can produce a high voltage. How does that happen, and what protection do the coils and the power supply need?",
   "group": "B",
   "group_title": "Before You Switch Anything On",
   "part": 2,
   "order": 3,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": true,
   "short_answer": "A magnet coil stores energy in its magnetic field, and that energy has to go somewhere when the current stops. The field is held up by the current and can only collapse by driving current onward. Open the circuit and the coil tries to keep the current flowing through the gap you just made. The voltage it generates to do so is whatever it takes. The tens of volts across the coil at steady current becomes kilovolts in a millisecond. That is enough to arc across a switch, puncture the coil insulation, or pass through a person holding the terminals. The protection is a deliberate place for that energy to go, at a voltage the coil, the supply and the people can tolerate. It has to be in place before the circuit is ever opened at current. Wouters’ 1949 rule states the function: never open the coil circuit at high current without a surge path across the coil ( Wouters (1949), p. 5 ). A surge path is a component across the coil’s terminals that conducts when the coil drives voltage across it, so the current keeps circulating through it instead of across the opened gap. For a first machine, the function list is the specification. A place for the energy to go, present before the first energization. A supply that can be ramped. A fault response chosen per fault. A coil proven to ground, covers, a grounded core. And the habit of never disconnecting anything with current flowing.",
   "short_answer_markdown": "A magnet coil stores energy in its magnetic field, and that energy has to go somewhere when the current stops. The field is held up by the current and can only collapse by driving current onward. Open the circuit and the coil tries to keep the current flowing through the gap you just made. The voltage it generates to do so is whatever it takes. The tens of volts across the coil at steady current becomes kilovolts in a millisecond. That is enough to arc across a switch, puncture the coil insulation, or pass through a person holding the terminals. The protection is a deliberate place for that energy to go, at a voltage the coil, the supply and the people can tolerate. It has to be in place before the circuit is ever opened at current. Wouters' 1949 rule states the function: never open the coil circuit at high current without a surge path across the coil (dg-218). A surge path is a component across the coil's terminals that conducts when the coil drives voltage across it, so the current keeps circulating through it instead of across the opened gap. For a first machine, the function list is the specification. A place for the energy to go, present before the first energization. A supply that can be ramped. A fault response chosen per fault. A coil proven to ground, covers, a grounded core. And the habit of never disconnecting anything with current flowing.",
   "sources": [
    "dg-036",
    "dg-057",
    "dg-073",
    "dg-087",
    "dg-1616",
    "dg-1670",
    "dg-1704",
    "dg-202",
    "dg-218"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/safety/#hazard-electrical",
    "https://cyclotroninfo.com/learn/safety/#hazard-magnet",
    "https://cyclotroninfo.com/learn/magnet-design/#coils",
    "https://cyclotroninfo.com/design-guide/coils/",
    "https://cyclotroninfo.com/design-guide/safety/"
   ],
   "refs": [
    "ground-what-floats",
    "safety-before-testing"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-switch-anything-on/#magnet-switch-off-voltage"
  },
  {
   "id": "ground-what-floats",
   "question": "What should be grounded and what needs to float? I have a filament heater supply, a filament bias supply, RF drive, a biased collector plate going to an electrometer, a magnet supply, vacuum-gauge supplies, and exposed metal everywhere. Which parts can be tied together, which must be isolated, and what does \"floating\" actually mean?",
   "group": "B",
   "group_title": "Before You Switch Anything On",
   "part": 2,
   "order": 4,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": true,
   "short_answer": "“Floating” always means floating relative to something; only what must sit at another potential floats; four ideas are easy to mix together. Protective earth is everything a hand can reach, chamber and magnet iron included, bonded to earth so that a fault has a low-impedance path and the protective devices operate. It does not promise that a fault can never produce a dangerous touch voltage, which is why isolation is proven before hands go in. It is never removed. A fault is a live conductor touching metal it should not; bonding gives the current a controlled return and holds reachable metal at one voltage, and a first fault from an isolated or current-limited supply may not open a fuse, so disconnection is designed per source. Touch voltage is the voltage between two things a person can touch at once. The RF return is the engineered low-impedance path the dee’s megahertz current flows back along. The measurement reference is the signal-low of a picoammeter, the question’s electrometer, where the measured current leaves, and the shield round its lead. Insulation leaks a little, and nearby conductors pass alternating current like a small capacitor. A floating circuit has no direct DC connection to a specified reference; leakage and stray capacitance still connect it electrically, which is why “floating” is never “disconnected”. A biased circuit is held at a chosen potential, ground-referenced or floating; both are statements about the fourth idea, never the first. Each box’s manual says which idea each terminal belongs to.",
   "short_answer_markdown": "\"Floating\" always means floating relative to something; only what must sit at another potential floats; four ideas are easy to mix together. **Protective earth** is everything a hand can reach, chamber and magnet iron included, bonded to earth so that a fault has a low-impedance path and the protective devices operate. It does not promise that a fault can never produce a dangerous touch voltage, which is why isolation is proven before hands go in. It is never removed. A fault is a live conductor touching metal it should not; bonding gives the current a controlled return and holds reachable metal at one voltage, and a first fault from an isolated or current-limited supply may not open a fuse, so disconnection is designed per source. Touch voltage is the voltage between two things a person can touch at once. **The RF return** is the engineered low-impedance path the dee's megahertz current flows back along. **The measurement reference** is the signal-low of a picoammeter, the question's electrometer, where the measured current leaves, and the shield round its lead. Insulation leaks a little, and nearby conductors pass alternating current like a small capacitor. **A floating circuit** has no direct DC connection to a specified reference; leakage and stray capacitance still connect it electrically, which is why \"floating\" is never \"disconnected\". A **biased** circuit is held at a chosen potential, ground-referenced or floating; both are statements about the fourth idea, never the first. Each box's manual says which idea each terminal belongs to.",
   "sources": [
    "dg-073",
    "dg-1616",
    "dg-1737",
    "dg-1864",
    "dg-202",
    "dg-240",
    "dg-241",
    "dg-282",
    "dg-310",
    "dg-317",
    "dg-516",
    "dg-715",
    "dg-888"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/learn/beam-measurement/#faraday-cup",
    "https://cyclotroninfo.com/learn/dee-coupling/#loop-or-probe",
    "https://cyclotroninfo.com/learn/ion-sources/#filament-sources",
    "https://cyclotroninfo.com/learn/safety/#hazard-electrical",
    "https://cyclotroninfo.com/library/"
   ],
   "refs": [
    "ground-loops",
    "ops-would-my-detector-see-it",
    "source-is-it-making-ions"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-switch-anything-on/#ground-what-floats"
  },
  {
   "id": "ground-loops",
   "question": "What are ground loops? How do I know if I have one, and how do I avoid them in my cyclotron?",
   "group": "B",
   "group_title": "Before You Switch Anything On",
   "part": 2,
   "order": 5,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": true,
   "short_answer": "A ground loop is a closed conductive path that runs through more than one ground or reference connection. Two instruments each earthed through their own mains cord, joined by a cable whose shield (the metal braid round the signal wire) is bonded at both ends, form one: cord, shield, cord, building wiring, back to cord. The two earth points are never at exactly the same potential, and the mains fields in the room thread the loop. A potential difference round a closed conductive path drives a current, and so does a changing magnetic field through the loop; the mains wiring’s magnetic fields alternate at line frequency, the mains frequency. Current circulates round it, and some of it flows in your signal return, the path the measured current takes back to the meter’s low terminal. That terminal is the signal reference, or signal low. On a picoammeter (the small-current meter) the result is hum (ripple at line frequency), an offset, or a reading that changes when another instrument is plugged in. The cure is to control where the measurement’s signal reference meets earth. It is never to remove a protective-earth connection, and those are different things even though the word “ground” covers both. For the loop through two instrument earths, one intentional DC bond of the signal reference removes the second earth connection; other reference loops, and AC paths closed through capacitance, still have to be considered. For a first beam that means one chosen point where signal low meets earth.",
   "short_answer_markdown": "A ground loop is a closed conductive path that runs through more than one ground or reference connection. Two instruments each earthed through their own mains cord, joined by a cable whose shield (the metal braid round the signal wire) is bonded at both ends, form one: cord, shield, cord, building wiring, back to cord. The two earth points are never at exactly the same potential, and the mains fields in the room thread the loop. A potential difference round a closed conductive path drives a current, and so does a changing magnetic field through the loop; the mains wiring's magnetic fields alternate at line frequency, the mains frequency. Current circulates round it, and some of it flows in your signal return, the path the measured current takes back to the meter's low terminal. That terminal is the signal reference, or signal low. On a picoammeter (the small-current meter) the result is hum (ripple at line frequency), an offset, or a reading that changes when another instrument is plugged in. The cure is to control where the measurement's signal reference meets earth. It is never to remove a protective-earth connection, and those are different things even though the word \"ground\" covers both. For the loop through two instrument earths, one intentional DC bond of the signal reference removes the second earth connection; other reference loops, and AC paths closed through capacitance, still have to be considered. For a first beam that means one chosen point where signal low meets earth.",
   "sources": [
    "dg-067",
    "dg-241",
    "dg-282",
    "dg-317",
    "dg-516"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/library/",
    "https://cyclotroninfo.com/learn/safety/#hazard-electrical",
    "https://cyclotroninfo.com/learn/safety/#hazard-rf"
   ],
   "refs": [
    "ground-what-floats",
    "ops-worked-yesterday",
    "ops-would-my-detector-see-it"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/before-you-switch-anything-on/#ground-loops"
  },
  {
   "id": "magnet-how-good",
   "question": "The books make the magnet requirements sound beyond amateur reach: field uniformity, steel homogeneity, and so on. Some amateur builds look much simpler. How good does my magnet really need to be?",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 1,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Good enough to find a beam is a lower bar than the books describe. The documented small machines that found beams had three things in common. Their magnetic field could be set and held steady. The resonance, the field at which the ions circle in step with the dee’s RF voltage and beam reaches the collector as a current peak, is narrow. Their field fell gently with radius, which flat iron poles often provide. That slight weakening gives an ion drifting up or down a small push back toward the middle of the gap, which keeps it off the dee. And their collector sat well inside the pole edge, where the field was still well behaved. The Rutgers 9-inch prototype made ten nanoamperes of beam current on flat pole faces with no shaping of the magnetic field ( Koeth (2006), p. 1 ). The professional tolerances you have read belong to machines that extracted beams at full radius for years. Those tolerances are a tenth of a percent variation of the field around a circle and parallelism to a part in fifty thousand. You may need them later.",
   "short_answer_markdown": "Good enough to find a beam is a lower bar than the books describe. The documented small machines that found beams had three things in common. Their magnetic field could be set and held steady. The resonance, the field at which the ions circle in step with the dee's RF voltage and beam reaches the collector as a current peak, is narrow. Their field fell gently with radius, which flat iron poles often provide. That slight weakening gives an ion drifting up or down a small push back toward the middle of the gap, which keeps it off the dee. And their collector sat well inside the pole edge, where the field was still well behaved. The Rutgers 9-inch prototype made ten nanoamperes of beam current on flat pole faces with no shaping of the magnetic field (dg-1727). The professional tolerances you have read belong to machines that extracted beams at full radius for years. Those tolerances are a tenth of a percent variation of the field around a circle and parallelism to a part in fifty thousand. You may need them later.",
   "sources": [
    "dg-003",
    "dg-009",
    "dg-012",
    "dg-027",
    "dg-031",
    "dg-090",
    "dg-096",
    "dg-116",
    "dg-132",
    "dg-143",
    "dg-148",
    "dg-152",
    "dg-1727",
    "dg-702"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/magnet-design/#field-spec",
    "https://cyclotroninfo.com/learn/magnet-design/#shimming",
    "https://cyclotroninfo.com/learn/beam-dynamics/#tune-diagram",
    "https://cyclotroninfo.com/builds/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-what-stops-them-hitting-the-dee",
    "lit-which-rules-apply",
    "magnet-dc-harmonics",
    "magnet-what-to-measure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#magnet-how-good"
  },
  {
   "id": "beam-what-stops-them-hitting-the-dee",
   "question": "The drawings show a flat spiral. What keeps the particles from wandering up or down and hitting the dee? Do I need \"focusing\"?",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 2,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Nothing in a perfectly uniform magnetic field does; the beam survives because a real field is slightly weaker at larger radius. The magnetic force on a moving charge acts at right angles to the field. Between flat poles the field points up or down, so nothing pushes an ion back toward the midplane, the plane halfway between the pole faces. An ion that starts a hair above it, or leaves the source tilted slightly upward, keeps that height or tilt forever. The dee is only a couple of centimetres tall. Where the field weakens with radius, the field lines bow outward. Where they bow, the field has a small sideways part, and that part’s push on the circling ion is vertical. An ion above the midplane feels a small push back down; one below, a push up. The ion wobbles gently about the midplane as it spirals; that wobble is what “focusing” means here. You need enough confinement and clearance for ions to survive to the collector. A flat-poled iron magnet often provides some of it, since its field weakens toward the edge. A detailed focusing analysis is not necessary for a first-beam demonstration.",
   "short_answer_markdown": "Nothing in a perfectly uniform magnetic field does; the beam survives because a real field is slightly weaker at larger radius. The magnetic force on a moving charge acts at right angles to the field. Between flat poles the field points up or down, so nothing pushes an ion back toward the midplane, the plane halfway between the pole faces. An ion that starts a hair above it, or leaves the source tilted slightly upward, keeps that height or tilt forever. The dee is only a couple of centimetres tall. Where the field weakens with radius, the field lines bow outward. Where they bow, the field has a small sideways part, and that part's push on the circling ion is vertical. An ion above the midplane feels a small push back down; one below, a push up. The ion wobbles gently about the midplane as it spirals; that wobble is what \"focusing\" means here. You need enough confinement and clearance for ions to survive to the collector. A flat-poled iron magnet often provides some of it, since its field weakens toward the edge. A detailed focusing analysis is not necessary for a first-beam demonstration.",
   "sources": [
    "dg-003",
    "dg-035",
    "dg-116",
    "dg-143",
    "dg-152",
    "dg-1637",
    "dg-1642",
    "dg-1727",
    "dg-1730",
    "dg-1796",
    "dg-1835",
    "dg-245",
    "dg-348",
    "dg-567",
    "dg-706",
    "dg-788"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#weak-focusing",
    "https://cyclotroninfo.com/learn/beam-dynamics/#two-numbers",
    "https://cyclotroninfo.com/learn/beam-dynamics/#recreate-1947",
    "https://cyclotroninfo.com/learn/ion-sources/#first-gap",
    "https://cyclotroninfo.com/learn/exercises/where-ions-are-born/",
    "https://cyclotroninfo.com/design-guide/beam-dynamics/"
   ],
   "refs": [
    "magnet-alignment",
    "magnet-dc-harmonics",
    "magnet-how-good",
    "source-where-does-it-go"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#beam-what-stops-them-hitting-the-dee"
  },
  {
   "id": "magnet-what-to-measure",
   "question": "Do I really need to map my magnet's field accurately? Is one reading in the middle enough, and which value goes into the frequency and energy calculations?",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 3,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Not necessarily, if your first goal is to demonstrate a beam. An ion circles at a rate set by the field’s strength (stronger means faster), its charge and its mass. The RF voltage must swing at that rate to push it at every gap crossing; the match is called resonance. For one ion at one RF frequency a beam therefore appears at one field, and the relation is simple: a proton resonates at 15.2 MHz per tesla, so 3.258 MHz calls for 214 mT. An inexpensive hand-held field meter (a Hall-effect teslameter, reading in millitesla) tells you approximately what field your magnet makes, and so roughly where to expect a peak. You do not need the field to four decimal places to look for that peak. Vary the field through the expected region and watch for a reproducible peak; it locates the field to run at far more precisely than the meter does. With a cheap meter the Mullins cyclotron found its proton peak at 214 mT for 3.258 MHz; the relation predicts 214.3 mT (build log, “It works!”). A detailed map becomes useful if no beam appears, if it disappears at larger radius, or to understand and improve the focusing, what keeps ions from drifting up or down into the dee.",
   "short_answer_markdown": "Not necessarily, if your first goal is to demonstrate a beam. An ion circles at a rate set by the field's strength (stronger means faster), its charge and its mass. The RF voltage must swing at that rate to push it at every gap crossing; the match is called resonance. For one ion at one RF frequency a beam therefore appears at one field, and the relation is simple: a proton resonates at 15.2 MHz per tesla, so 3.258 MHz calls for 214 mT. An inexpensive hand-held field meter (a Hall-effect teslameter, reading in millitesla) tells you approximately what field your magnet makes, and so roughly where to expect a peak. You do not need the field to four decimal places to look for that peak. Vary the field through the expected region and watch for a reproducible peak; it locates the field to run at far more precisely than the meter does. With a cheap meter the Mullins cyclotron found its proton peak at 214 mT for 3.258 MHz; the relation predicts 214.3 mT (build log, \"It works!\"). A detailed map becomes useful if no beam appears, if it disappears at larger radius, or to understand and improve the focusing, what keeps ions from drifting up or down into the dee.",
   "sources": [
    "dg-002",
    "dg-013",
    "dg-014",
    "dg-015",
    "dg-017",
    "dg-028",
    "dg-090",
    "dg-104",
    "dg-105",
    "dg-141",
    "dg-147",
    "dg-1731",
    "dg-180",
    "dg-243",
    "dg-511",
    "dg-980"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/magnet-design/#measurement",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/beam-measurement/#energy",
    "https://cyclotroninfo.com/library/amateur/#mullins",
    "https://cyclotroninfo.com/builds/",
    "https://cyclotroninfo.com/calculators/cyclotron-energy/"
   ],
   "refs": [
    "magnet-dc-harmonics",
    "magnet-how-good",
    "ops-finding-first-beam",
    "ops-worked-yesterday"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#magnet-what-to-measure"
  },
  {
   "id": "magnet-dc-harmonics",
   "question": "How can my magnet's field have \"harmonics\" when the magnet runs on DC? I thought harmonics were an audio thing, like musical notes. Do I have to worry about them?",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 4,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "The harmonics are in space, and for a first beam most builders need not decompose anything. Walk once around a circle in the gap at fixed radius, reading the magnetic field as you go. A perfect magnet reads the same everywhere. A real one reads a little higher on one side than the other, or a little higher at two opposite places. That pattern around the circle can be broken into sine waves of one cycle per lap, two per lap, and so on, as a sound is broken into its notes. The first harmonic is the once-around component. To the ion the distinction between space and time disappears, because it goes round that circle millions of times a second. A once-around bump in space is a push it feels once per turn. A push repeated in step with the motion can add up. Hence the word resonance, as in the RF. If the beam appears and behaves, the imperfections were tolerable for that result, which is not the same as saying they were small. The question returns if the beam only appears with the source off-center or dies at radius.",
   "short_answer_markdown": "The harmonics are in space, and for a first beam most builders need not decompose anything. Walk once around a circle in the gap at fixed radius, reading the magnetic field as you go. A perfect magnet reads the same everywhere. A real one reads a little higher on one side than the other, or a little higher at two opposite places. That pattern around the circle can be broken into sine waves of one cycle per lap, two per lap, and so on, as a sound is broken into its notes. The first harmonic is the once-around component. To the ion the distinction between space and time disappears, because it goes round that circle millions of times a second. A once-around bump in space is a push it feels once per turn. A push repeated in step with the motion can add up. Hence the word resonance, as in the RF. If the beam appears and behaves, the imperfections were tolerable for that result, which is not the same as saying they were small. The question returns if the beam only appears with the source off-center or dies at radius.",
   "sources": [
    "dg-012",
    "dg-015",
    "dg-050",
    "dg-1873",
    "dg-342",
    "dg-506",
    "dg-587",
    "dg-696"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-dynamics/#imperfection-lab",
    "https://cyclotroninfo.com/learn/beam-dynamics/#tune-diagram",
    "https://cyclotroninfo.com/learn/magnet-design/#shimming",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/design-guide/beam-dynamics/"
   ],
   "refs": [
    "beam-which-ion",
    "magnet-how-good",
    "magnet-what-to-measure",
    "rf-which-resonance"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#magnet-dc-harmonics"
  },
  {
   "id": "magnet-alignment",
   "question": "How accurately do I need to line up the magnet, chamber, dee, ion source and collector? Which dimensions are really critical, and what should be adjustable?",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 5,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Start with fit, clearance and reproducibility, and bring in the ten-thousandths of an inch later. For a first beam the dee must clear the lids, the chamber’s top and bottom plates, with margin all round. The dee carries the RF voltage, and a close lid shortens the distance a spark has to jump. Ions stay near the height they start at, so the source must sit at mid-height in the gap. The collector must sweep through the radii where beam is expected. The chamber must go back into the magnet the same way every time, so a setting found one day means something the next. The source position and the collector radius are tuned for beam rather than measured to a drawing. Several documented machines made them adjustable from outside the vacuum, a convenience worth copying rather than a requirement. A gap narrower on one side makes the magnetic field stronger on that side, and a field stronger on one side pushes the orbit off center. The professional precision, gap variation in thousandths and parallelism in parts per fifty thousand, keeps that once-around error small and the orbit centered. Pursue it if the beam’s behaviour points there.",
   "short_answer_markdown": "Start with fit, clearance and reproducibility, and bring in the ten-thousandths of an inch later. For a first beam the dee must clear the lids, the chamber's top and bottom plates, with margin all round. The dee carries the RF voltage, and a close lid shortens the distance a spark has to jump. Ions stay near the height they start at, so the source must sit at mid-height in the gap. The collector must sweep through the radii where beam is expected. The chamber must go back into the magnet the same way every time, so a setting found one day means something the next. The source position and the collector radius are tuned for beam rather than measured to a drawing. Several documented machines made them adjustable from outside the vacuum, a convenience worth copying rather than a requirement. A gap narrower on one side makes the magnetic field stronger on that side, and a field stronger on one side pushes the orbit off center. The professional precision, gap variation in thousandths and parallelism in parts per fifty thousand, keeps that once-around error small and the orbit centered. Pursue it if the beam's behaviour points there.",
   "sources": [
    "dg-006",
    "dg-013",
    "dg-020",
    "dg-029",
    "dg-1149",
    "dg-1325",
    "dg-1381",
    "dg-150",
    "dg-151",
    "dg-1655",
    "dg-1852",
    "dg-2058",
    "dg-244",
    "dg-283",
    "dg-293",
    "dg-640",
    "dg-645",
    "dg-788",
    "dg-789"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/magnet-design/#forces-tolerances",
    "https://cyclotroninfo.com/learn/magnet-design/#measurement",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#lids",
    "https://cyclotroninfo.com/learn/ion-sources/#first-gap",
    "https://cyclotroninfo.com/calculators/lid-deflection/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "magnet-dc-harmonics",
    "magnet-what-to-measure",
    "rf-bench-vs-installed",
    "source-where-does-it-go"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#magnet-alignment"
  },
  {
   "id": "magnet-permanent-magnet",
   "question": "Can I use a permanent magnet instead of an electromagnet? It would save the power supply, the cooling and the coils.",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 6,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "You can, and it removes what you list: ampere-turns (coil current times turns), watts, cooling water, the regulated supply and its interlocks (automatic shut-offs). It also removes the adjustment most first beams are found with. The usual way to bring a classical cyclotron into resonance is a sweep: the RF frequency held fixed, the magnet current run through a range. The field is swept because it is cheap to vary and the RF is not ( Livingston Blewett (1962), p. 156 ). The RF is harder to vary because the dee gets its voltage from a tuned circuit that builds voltage only near its own resonant frequency ( I keep seeing cyclotron resonance , RF resonance and impedance matching . Are these the same thing? What has to match what? ). Without the current sweep, the sweep has to be done with the RF frequency, and the field has to stay put against room temperature well enough that the RF can follow it. A professional group solved both at exactly this scale. In the amateur record one machine was built this way and has published no performance. For a first machine, choose the permanent magnet if you have an RF system that can deliver dee voltage across the range you will need to sweep, and a plan for temperature. Choose the electromagnet if you want the current sweep.",
   "short_answer_markdown": "You can, and it removes what you list: ampere-turns (coil current times turns), watts, cooling water, the regulated supply and its interlocks (automatic shut-offs). It also removes the adjustment most first beams are found with. The usual way to bring a classical cyclotron into resonance is a sweep: the RF frequency held fixed, the magnet current run through a range. The field is swept because it is cheap to vary and the RF is not (dg-243). The RF is harder to vary because the dee gets its voltage from a tuned circuit that builds voltage only near its own resonant frequency (rf-which-resonance). Without the current sweep, the sweep has to be done with the RF frequency, and the field has to stay put against room temperature well enough that the RF can follow it. A professional group solved both at exactly this scale. In the amateur record one machine was built this way and has published no performance. For a first machine, choose the permanent magnet if you have an RF system that can deliver dee voltage across the range you will need to sweep, and a plan for temperature. Choose the electromagnet if you want the current sweep.",
   "sources": [
    "dg-047",
    "dg-1442",
    "dg-1549",
    "dg-1552",
    "dg-1571",
    "dg-2108",
    "dg-2109",
    "dg-2111",
    "dg-2116",
    "dg-2122",
    "dg-2123",
    "dg-2130",
    "dg-2138",
    "dg-2141",
    "dg-2144",
    "dg-2145",
    "dg-2146",
    "dg-2148",
    "dg-2149",
    "dg-2151",
    "dg-243"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/magnet-design/#permanent-magnets",
    "https://cyclotroninfo.com/learn/safety/#hazard-magnet",
    "https://cyclotroninfo.com/design-guide/magnet/",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "magnet-how-good",
    "ops-finding-first-beam",
    "rf-which-resonance"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#magnet-permanent-magnet"
  },
  {
   "id": "magnet-surplus-mri-nmr",
   "question": "Can I build around a surplus magnet, like an old NMR or MRI magnet, or a stack of microwave-oven transformer cores?",
   "group": "C",
   "group_title": "The Magnet, and Staying in the Gap",
   "part": 2,
   "order": 7,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "An iron-pole laboratory electromagnet, the kind sold with older NMR and EPR instruments, is the most common successful route in the census. It is already a cyclotron magnet in all but pole shape: iron poles across a gap, a coil current to set and vary the field. A clinical MRI magnet, and the superconducting solenoids of high-resolution NMR systems, are a different kind of object and an impractical start for a small classical cyclotron. Each is a coil round an open bore, run cold, with its field along the bore, no pole faces to shape and no convenient way to vary the field while commissioning. Transformer cores are the wrong size: a cyclotron pole needs dozens of times their gap area and iron return path. The trap with the laboratory magnet is designing to its headline field rather than to its gap and usable radius, which ends inside the pole edge. Two tesla implies a 30 MHz RF system on a small orbit. The ions’ resonance frequency rises with the field (protons, 15.2 MHz per tesla; the cyclotron-energy calculator does the sum), and a stronger field bends them tighter. Measure the magnet you have and design the rest to it.",
   "short_answer_markdown": "An iron-pole laboratory electromagnet, the kind sold with older NMR and EPR instruments, is the most common successful route in the census. It is already a cyclotron magnet in all but pole shape: iron poles across a gap, a coil current to set and vary the field. A clinical MRI magnet, and the superconducting solenoids of high-resolution NMR systems, are a different kind of object and an impractical start for a small classical cyclotron. Each is a coil round an open bore, run cold, with its field along the bore, no pole faces to shape and no convenient way to vary the field while commissioning. Transformer cores are the wrong size: a cyclotron pole needs dozens of times their gap area and iron return path. The trap with the laboratory magnet is designing to its headline field rather than to its gap and usable radius, which ends inside the pole edge. Two tesla implies a 30 MHz RF system on a small orbit. The ions' resonance frequency rises with the field (protons, 15.2 MHz per tesla; the cyclotron-energy calculator does the sum), and a stronger field bends them tighter. Measure the magnet you have and design the rest to it.",
   "sources": [
    "dg-017",
    "dg-022",
    "dg-023",
    "dg-071",
    "dg-116",
    "dg-121",
    "dg-128",
    "dg-130",
    "dg-132",
    "dg-134",
    "dg-146",
    "dg-148",
    "dg-150",
    "dg-960"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/magnet-design/#iron-circuit",
    "https://cyclotroninfo.com/learn/magnet-design/#gap-and-poles",
    "https://cyclotroninfo.com/learn/choosing-your-machine/#design-point",
    "https://cyclotroninfo.com/builds/",
    "https://cyclotroninfo.com/design-guide/magnet/"
   ],
   "refs": [
    "goal-copy-and-swap",
    "goal-how-it-fits-together",
    "magnet-how-good",
    "magnet-permanent-magnet"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-magnet/#magnet-surplus-mri-nmr"
  },
  {
   "id": "source-is-it-making-ions",
   "question": "My filament is glowing. How do I know it's actually making ions? And what are the filament current, the bias voltage and the emission current each telling me?",
   "group": "D",
   "group_title": "Making Ions, and the First Few Turns",
   "part": 2,
   "order": 1,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "A glow proves heat. Three currents are involved, usually read on three different meters. Heater current, in amperes, flows through the filament wire itself and only warms it. Emission current, usually milliamperes, is the electrons that leave the hot wire under bias. Bias is a voltage holding the filament negative with respect to the metal around it; it pulls the freed electrons away, and those landing on grounded metal flow back to the supply through its return lead, where the emission meter counts them. Beam current, nanoamperes to microamperes, is what the collector eventually sees. A bright filament can be emitting almost nothing, because emission turns on sharply with temperature. So the first check is the emission meter with no gas in the chamber. With gas present the electrons sustain a discharge in it, which the emission meter also reads, so the second check is whether emission rises when gas is admitted. The rise is consistent with ionization and does not prove it; neither check proves that ions are reaching the accelerating gap at the dee edge. For a first beam, set emission on the meter rather than by the glow and look for that rise ( Where exactly should the filament or the source opening go? Is at the center good enough, and how do the first orbits avoid hitting the source? ).",
   "short_answer_markdown": "A glow proves heat. Three currents are involved, usually read on three different meters. Heater current, in amperes, flows through the filament wire itself and only warms it. Emission current, usually milliamperes, is the electrons that leave the hot wire under bias. Bias is a voltage holding the filament negative with respect to the metal around it; it pulls the freed electrons away, and those landing on grounded metal flow back to the supply through its return lead, where the emission meter counts them. Beam current, nanoamperes to microamperes, is what the collector eventually sees. A bright filament can be emitting almost nothing, because emission turns on sharply with temperature. So the first check is the emission meter with no gas in the chamber. With gas present the electrons sustain a discharge in it, which the emission meter also reads, so the second check is whether emission rises when gas is admitted. The rise is consistent with ionization and does not prove it; neither check proves that ions are reaching the accelerating gap at the dee edge. For a first beam, set emission on the meter rather than by the glow and look for that rise (source-where-does-it-go).",
   "sources": [
    "dg-1414",
    "dg-1737",
    "dg-1814",
    "dg-1866",
    "dg-1867",
    "dg-377",
    "dg-379",
    "dg-391",
    "dg-402",
    "dg-607"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/ion-sources/#filament-sources",
    "https://cyclotroninfo.com/learn/ion-sources/#arc-behavior",
    "https://cyclotroninfo.com/learn/ion-sources/#first-gap",
    "https://cyclotroninfo.com/learn/exercises/where-ions-are-born/",
    "https://cyclotroninfo.com/design-guide/ion-source/"
   ],
   "refs": [
    "beam-is-it-working",
    "source-timing-with-rf",
    "source-where-does-it-go",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/making-ions/#source-is-it-making-ions"
  },
  {
   "id": "source-where-does-it-go",
   "question": "Where exactly should the filament or the source opening go? Is \"at the center\" good enough, and how do the first orbits avoid hitting the source?",
   "group": "D",
   "group_title": "Making Ions, and the First Few Turns",
   "part": 2,
   "order": 2,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Near the center of the machine, halfway between the magnet’s pole faces, the iron above and below the chamber. “At the center” is good enough to start. The requirement is that ions can get from where they are made into the accelerating gap without immediately striking nearby metal. The first orbit misses the source because the voltage on the dee pulls the ion away as soon as it is made. The magnet’s field then bends the moving ion round, and a faster ion makes a wider circle. One crossing of the gap at a few hundred volts gives a proton a first-turn radius of a few millimetres in a tesla-class magnetic field. The ion passes the source on the far side of that circle, is pulled again, and moves out on its spiral. A bare filament or a small button is cleared easily. A source housing the size of a thumb is not. The Mullins machine’s source is a bare tungsten hairpin on a ceramic button, with no housing of any kind (hosted rf-dee-source note, §6). It found beam. Put the source on the axis of the poles and let the beam tell you whether to move it.",
   "short_answer_markdown": "Near the center of the machine, halfway between the magnet's pole faces, the iron above and below the chamber. \"At the center\" is good enough to start. The requirement is that ions can get from where they are made into the accelerating gap without immediately striking nearby metal. The first orbit misses the source because the voltage on the dee pulls the ion away as soon as it is made. The magnet's field then bends the moving ion round, and a faster ion makes a wider circle. One crossing of the gap at a few hundred volts gives a proton a first-turn radius of a few millimetres in a tesla-class magnetic field. The ion passes the source on the far side of that circle, is pulled again, and moves out on its spiral. A bare filament or a small button is cleared easily. A source housing the size of a thumb is not. The Mullins machine's source is a bare tungsten hairpin on a ceramic button, with no housing of any kind (hosted rf-dee-source note, §6). It found beam. Put the source on the axis of the poles and let the beam tell you whether to move it.",
   "sources": [
    "dg-013",
    "dg-1386",
    "dg-1387",
    "dg-1522",
    "dg-1646",
    "dg-1651",
    "dg-1735",
    "dg-1840",
    "dg-1852",
    "dg-2058",
    "dg-2103",
    "dg-244",
    "dg-245",
    "dg-348",
    "dg-576",
    "dg-788",
    "dg-804"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/ion-sources/#the-job",
    "https://cyclotroninfo.com/learn/ion-sources/#first-gap",
    "https://cyclotroninfo.com/learn/exercises/where-ions-are-born/",
    "https://cyclotroninfo.com/calculators/path-length/",
    "https://cyclotroninfo.com/learn/beam-dynamics/#two-numbers",
    "https://cyclotroninfo.com/library/amateur/#mullins",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "beam-what-stops-them-hitting-the-dee",
    "magnet-alignment",
    "ops-beam-near-center-not-farther-out",
    "source-is-it-making-ions",
    "source-timing-with-rf"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/making-ions/#source-where-does-it-go"
  },
  {
   "id": "source-timing-with-rf",
   "question": "My ion source is on all the time, but the RF keeps reversing. How do the ions start at the right moment? Do I have to synchronize the source with the RF?",
   "group": "D",
   "group_title": "Making Ions, and the First Few Turns",
   "part": 2,
   "order": 3,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "No. You do not synchronize anything; the gap does the choosing. The source makes ions continuously. The RF voltage on the dee makes an electric field across the gap, a push on any charged particle, and the push reverses direction each half-cycle with the voltage. The ions are positive, so an ion is pushed the way the field points at the instant it forms. An ion born while the field points out of the source is pulled out, crosses the gap, and the magnet’s field brings it round for its next push. An ion born while the field points into the source is pushed back into it or never leaves. For a typical internal source, one sitting inside the vacuum at the machine’s centre, roughly a third of each RF cycle is usable. The fraction depends on the geometry and is a ceiling rather than a measured capture efficiency ( Smirnov (2021), p. 56 ). A hot filament could not be switched at megahertz rates in any case; it heats and cools far too slowly. Your part is the match between the magnet’s field and the RF frequency, which decides whether an accepted ion comes back to the gap in step on every turn.",
   "short_answer_markdown": "No. You do not synchronize anything; the gap does the choosing. The source makes ions continuously. The RF voltage on the dee makes an electric field across the gap, a push on any charged particle, and the push reverses direction each half-cycle with the voltage. The ions are positive, so an ion is pushed the way the field points at the instant it forms. An ion born while the field points out of the source is pulled out, crosses the gap, and the magnet's field brings it round for its next push. An ion born while the field points into the source is pushed back into it or never leaves. For a typical internal source, one sitting inside the vacuum at the machine's centre, roughly a third of each RF cycle is usable. The fraction depends on the geometry and is a ceiling rather than a measured capture efficiency (dg-576). A hot filament could not be switched at megahertz rates in any case; it heats and cools far too slowly. Your part is the match between the magnet's field and the RF frequency, which decides whether an accepted ion comes back to the gap in step on every turn.",
   "sources": [
    "dg-1591",
    "dg-1809",
    "dg-245",
    "dg-246",
    "dg-309",
    "dg-348",
    "dg-402",
    "dg-507",
    "dg-576",
    "dg-596",
    "dg-705",
    "dg-889"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#phase-stability",
    "https://cyclotroninfo.com/learn/ion-sources/#the-job",
    "https://cyclotroninfo.com/learn/exercises/out-of-step/",
    "https://cyclotroninfo.com/learn/beam-quality/#energy-spread",
    "https://cyclotroninfo.com/learn/beam-quality/#tradeoff",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "magnet-what-to-measure",
    "ops-beam-near-center-not-farther-out",
    "rf-which-resonance",
    "source-how-many-turns",
    "source-where-does-it-go"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/making-ions/#source-timing-with-rf"
  },
  {
   "id": "source-how-many-turns",
   "question": "How many turns should a proton make before it reaches my target energy? Fewer turns sounds easier, but isn't the whole point of a cyclotron to accelerate over lots of turns?",
   "group": "D",
   "group_title": "Making Ions, and the First Few Turns",
   "part": 2,
   "order": 4,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "You do not choose the turn count. It follows from the energy you aim for and the dee voltage, and a smaller count is generally the easier one. For a proton in a nearly uniform magnetic field, the energy at a given radius is set mainly by the field and that radius. The dee voltage plays no part in it ( Prechtl Wolf (2020), p. 31 ; the energy calculator). That holds for a known species on a centred orbit, well below the speed of light. A singly charged ion gains one electron-volt for each volt it crosses. Each turn crosses the gap twice, so with ideal timing an ion gains twice the gap voltage in electron-volts per turn. The turn count is the final energy divided by that (the path-length calculator). In the site’s simulated laboratory, halving the dee voltage doubles the turn count and leaves the collected energy where it was (guided exercise “More voltage, same energy”). The cyclotron’s point is the reuse of one modest voltage in place of a very large one; the many turns are a consequence of that reuse. Run the voltage you can make, then read the turn count off the calculator as the trip your ions must survive.",
   "short_answer_markdown": "You do not choose the turn count. It follows from the energy you aim for and the dee voltage, and a smaller count is generally the easier one. For a proton in a nearly uniform magnetic field, the energy at a given radius is set mainly by the field and that radius. The dee voltage plays no part in it (dg-1376; the energy calculator). That holds for a known species on a centred orbit, well below the speed of light. A singly charged ion gains one electron-volt for each volt it crosses. Each turn crosses the gap twice, so with ideal timing an ion gains twice the gap voltage in electron-volts per turn. The turn count is the final energy divided by that (the path-length calculator). In the site's simulated laboratory, halving the dee voltage doubles the turn count and leaves the collected energy where it was (guided exercise \"More voltage, same energy\"). The cyclotron's point is the reuse of one modest voltage in place of a very large one; the many turns are a consequence of that reuse. Run the voltage you can make, then read the turn count off the calculator as the trip your ions must survive.",
   "sources": [
    "dg-1376",
    "dg-1398",
    "dg-1508",
    "dg-1517",
    "dg-1531",
    "dg-1716",
    "dg-1736",
    "dg-1786",
    "dg-1809",
    "dg-246",
    "dg-270",
    "dg-272",
    "dg-273",
    "dg-301",
    "dg-303",
    "dg-359",
    "dg-495",
    "dg-596"
   ],
   "links": [
    "https://cyclotroninfo.com/calculators/path-length/",
    "https://cyclotroninfo.com/calculators/cyclotron-energy/",
    "https://cyclotroninfo.com/learn/exercises/more-voltage-same-energy/",
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#dees-and-gap",
    "https://cyclotroninfo.com/learn/beam-quality/#energy-spread",
    "https://cyclotroninfo.com/learn/beam-extraction/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "ops-beam-near-center-not-farther-out",
    "rf-sub-kilovolt-dees",
    "rf-watts-to-volts",
    "source-timing-with-rf",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/making-ions/#source-how-many-turns"
  },
  {
   "id": "source-space-charge",
   "question": "What is \"space charge\", and when would it matter in a small cyclotron? Is it a problem in the source, during acceleration, or both?",
   "group": "D",
   "group_title": "Making Ions, and the First Few Turns",
   "part": 2,
   "order": 5,
   "tags": [
    "build"
   ],
   "version": 3,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Space charge is the beam pushing on itself. Every ion is positive, so the ions repel one another. With enough of them close together, that spreads the beam faster than the magnet’s focusing, the gentle push back toward the midplane described in group C, can gather it. With a small cyclotron’s weak beam, it usually matters much less in the spiral than ions starting in the wrong place, out of step with the RF, or hitting metal. There is no single current below which it can be dismissed. There is a scale, though. The design literature most often puts the point where space charge takes over at a few hundred microamperes circulating, for the class of machine it describes ( Smirnov (2021), p. 40 , Smirnov’s review). Documented amateur machines run picoamperes to nanoamperes at the collector (census). That gap puts space charge low on the first-beam list, as an expectation rather than a proof; look for the ordinary losses first. In the source, where the charges are far denser, it acts routinely. The bulk of the ionised gas stays close to electrically neutral. The effect sits in thin layers next to the electrodes and at the opening where ions are drawn out, and sets how much current can leave.",
   "short_answer_markdown": "Space charge is the beam pushing on itself. Every ion is positive, so the ions repel one another. With enough of them close together, that spreads the beam faster than the magnet's focusing, the gentle push back toward the midplane described in group C, can gather it. With a small cyclotron's weak beam, it usually matters much less in the spiral than ions starting in the wrong place, out of step with the RF, or hitting metal. There is no single current below which it can be dismissed. There is a scale, though. The design literature most often puts the point where space charge takes over at a few hundred microamperes circulating, for the class of machine it describes (dg-580, Smirnov's review). Documented amateur machines run picoamperes to nanoamperes at the collector (census). That gap puts space charge low on the first-beam list, as an expectation rather than a proof; look for the ordinary losses first. In the source, where the charges are far denser, it acts routinely. The bulk of the ionised gas stays close to electrically neutral. The effect sits in thin layers next to the electrodes and at the opening where ions are drawn out, and sets how much current can leave.",
   "sources": [
    "dg-1649",
    "dg-1900",
    "dg-2069",
    "dg-2076",
    "dg-375",
    "dg-376",
    "dg-420",
    "dg-507",
    "dg-580"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-quality/#tradeoff",
    "https://cyclotroninfo.com/learn/beam-quality/#good-beam",
    "https://cyclotroninfo.com/learn/ion-sources/#first-gap",
    "https://cyclotroninfo.com/learn/ion-sources/#slit-geometry",
    "https://cyclotroninfo.com/learn/ion-sources/#gas-feed",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-which-terms-can-wait",
    "source-is-it-making-ions",
    "source-timing-with-rf",
    "source-where-does-it-go",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/making-ions/#source-space-charge"
  },
  {
   "id": "vac-what-pressure",
   "question": "What pressure should I aim for when the cyclotron is running? Don't I need more gas to make ions, but less gas to let them spiral around?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 1,
   "tags": [
    "build",
    "run"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "There is no one right pressure, because the conflict you describe is real. The source needs enough gas to make ions; the accelerated ions need the gas gone, so they can finish their spiral without colliding with it. Between the two there is a usable window, and neither more gas nor a better vacuum must always give more beam. Below the window the source has too little gas to ionize. Above it the ions you made collide with gas on their way round and are lost, and the resonance peaks (the peaks in beam current as the magnetic field or the RF frequency is swept) grow broad and shift ( Yuly (2013), p. 5 ). Houghton’s record has the machine operating between about 10⁻⁶ and 10⁻⁴ torr, with the best current in a narrower band around 1 to 3 × 10⁻⁵ ( Yuly (2013), p. 5 , Fuller (2013), p. 51-52 ). Those are two theses on one machine, an operating window and an optimum inside it, not to be averaged. Because the window has two sides, the setting is found by trying: pump down with the gas off, admit gas, and change the gas alone while watching the collector current.",
   "short_answer_markdown": "There is no one right pressure, because the conflict you describe is real. The source needs enough gas to make ions; the accelerated ions need the gas gone, so they can finish their spiral without colliding with it. Between the two there is a usable window, and neither more gas nor a better vacuum must always give more beam. Below the window the source has too little gas to ionize. Above it the ions you made collide with gas on their way round and are lost, and the resonance peaks (the peaks in beam current as the magnetic field or the RF frequency is swept) grow broad and shift (dg-405). Houghton's record has the machine operating between about 10⁻⁶ and 10⁻⁴ torr, with the best current in a narrower band around 1 to 3 × 10⁻⁵ (dg-405, dg-403). Those are two theses on one machine, an operating window and an optimum inside it, not to be averaged. Because the window has two sides, the setting is found by trying: pump down with the gas off, admit gas, and change the gas alone while watching the collector current.",
   "sources": [
    "dg-1398",
    "dg-1509",
    "dg-1868",
    "dg-370",
    "dg-391",
    "dg-403",
    "dg-405",
    "dg-409",
    "dg-414",
    "dg-460",
    "dg-465"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/vacuum/#from-the-beam",
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/learn/vacuum/#worked-budget",
    "https://cyclotroninfo.com/learn/ion-sources/#gas-feed",
    "https://cyclotroninfo.com/calculators/vacuum/",
    "https://cyclotroninfo.com/learn/exercises/what-the-gas-costs/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-which-ion",
    "source-how-many-turns",
    "vac-can-i-trust-the-gauge",
    "vac-gas-control",
    "vac-what-pump"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-what-pressure"
  },
  {
   "id": "vac-what-pump",
   "question": "Do I need a diffusion pump or a turbopump, or will a good roughing pump do?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 2,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "A roughing pump alone will not get you there, and every machine in the census that made beam and reports its pumping had a second pump. A roughing pump takes a chamber from atmosphere to a rough vacuum; the common oil-sealed rotary-vane kind stops in the 10⁻² to 10⁻³ torr region. A cyclotron needs two to three decades below that, a decade being a factor of ten. At 10⁻³ torr a proton in hydrogen goes centimetres, less than one turn, before it pulls an electron off a gas molecule (that distance is its capture mean free path). As a neutral atom it no longer curves in the magnet’s field and flies into the wall. The second pump, between chamber and roughing pump, is a diffusion pump (boiling oil vapour sweeps the gas out) or a turbopump (a fast turbine knocks it out). Beam machines in the census used both and both reached the band; the record does not show one technology failing where the other succeeded. Most builders used whichever turned up surplus, donated or affordable. Each has housekeeping with a safety side. The roughing pump’s exhaust also carries an oil mist into the room, a health matter the safety page covers under pump exhaust.",
   "short_answer_markdown": "A roughing pump alone will not get you there, and every machine in the census that made beam and reports its pumping had a second pump. A roughing pump takes a chamber from atmosphere to a rough vacuum; the common oil-sealed rotary-vane kind stops in the 10⁻² to 10⁻³ torr region. A cyclotron needs two to three decades below that, a decade being a factor of ten. At 10⁻³ torr a proton in hydrogen goes centimetres, less than one turn, before it pulls an electron off a gas molecule (that distance is its capture mean free path). As a neutral atom it no longer curves in the magnet's field and flies into the wall. The second pump, between chamber and roughing pump, is a diffusion pump (boiling oil vapour sweeps the gas out) or a turbopump (a fast turbine knocks it out). Beam machines in the census used both and both reached the band; the record does not show one technology failing where the other succeeded. Most builders used whichever turned up surplus, donated or affordable. Each has housekeeping with a safety side. The roughing pump's exhaust also carries an oil mist into the room, a health matter the safety page covers under pump exhaust.",
   "sources": [
    "dg-1153",
    "dg-1154",
    "dg-1348",
    "dg-1403",
    "dg-1405",
    "dg-1832",
    "dg-1853",
    "dg-432",
    "dg-440",
    "dg-445",
    "dg-446",
    "dg-471",
    "dg-842",
    "dg-846",
    "dg-847"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/vacuum/#pumps",
    "https://cyclotroninfo.com/learn/vacuum/#effective-speed",
    "https://cyclotroninfo.com/learn/safety/",
    "https://cyclotroninfo.com/library/amateur/#mullins",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "safety-before-testing",
    "vac-how-clean",
    "vac-pump-rated-better",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-what-pump"
  },
  {
   "id": "vac-pump-rated-better",
   "question": "My pump is rated to reach a much better vacuum than I need. Why can't my chamber get there? What are pumping speed and conductance?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 3,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Because the rating describes the pump with nothing attached. The ultimate pressure on the nameplate is what the pump reaches blanked off, with its inlet closed. The pumping speed is how many litres per second it swallows at its inlet. The conductance is how many litres per second the pipe between pump and chamber lets through, and a pipe can only subtract. Gas keeps entering a working chamber, from the walls and the seals and, when it runs, from the source’s feed. The chamber settles where the gas arriving equals the gas being removed. Its pressure is roughly the base plus the admitted gas divided by the effective speed, if admitting gas does not change the rest. The base is the pressure the system reaches with no gas admitted, the walls and seals already counted; the effective speed is what the pump delivers after the pipe has taken its share. The Mullins machine’s three readings show where the gap opens: pump alone against empty chamber is the pipe and the bare chamber; empty against working is the source’s feed and the fitted parts. Its surplus diffusion pump alone reached 1.8 × 10⁻⁶ torr. The empty chamber on it reached 3.4 × 10⁻⁶. The working chamber sat near 10⁻⁵ after hours of pumping (hosted chamber-vacuum note, §12). The same three readings on your own system show where the pressure is lost.",
   "short_answer_markdown": "Because the rating describes the pump with nothing attached. The *ultimate pressure* on the nameplate is what the pump reaches blanked off, with its inlet closed. The *pumping speed* is how many litres per second it swallows at its inlet. The *conductance* is how many litres per second the pipe between pump and chamber lets through, and a pipe can only subtract. Gas keeps entering a working chamber, from the walls and the seals and, when it runs, from the source's feed. The chamber settles where the gas arriving equals the gas being removed. Its pressure is roughly the base plus the admitted gas divided by the effective speed, if admitting gas does not change the rest. The base is the pressure the system reaches with no gas admitted, the walls and seals already counted; the effective speed is what the pump delivers after the pipe has taken its share. The Mullins machine's three readings show where the gap opens: pump alone against empty chamber is the pipe and the bare chamber; empty against working is the source's feed and the fitted parts. Its surplus diffusion pump alone reached 1.8 × 10⁻⁶ torr. The empty chamber on it reached 3.4 × 10⁻⁶. The working chamber sat near 10⁻⁵ after hours of pumping (hosted chamber-vacuum note, §12). The same three readings on your own system show where the pressure is lost.",
   "sources": [
    "dg-1154",
    "dg-1347",
    "dg-1348",
    "dg-1401",
    "dg-1402",
    "dg-356",
    "dg-370",
    "dg-432",
    "dg-440",
    "dg-446",
    "dg-447",
    "dg-450",
    "dg-451",
    "dg-453",
    "dg-456",
    "dg-458",
    "dg-467",
    "dg-649",
    "dg-650",
    "dg-837",
    "dg-840",
    "dg-842",
    "dg-844"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/vacuum/#effective-speed",
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/learn/vacuum/#seals-and-leaks",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#welds",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#elastomers",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "vac-can-i-trust-the-gauge",
    "vac-how-clean",
    "vac-what-materials",
    "vac-what-pump"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-pump-rated-better"
  },
  {
   "id": "vac-can-i-trust-the-gauge",
   "question": "Can I trust the pressure my vacuum gauge shows? Is that the pressure where the particles are travelling?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 4,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Yes, as a repeatable number for finding a setting and returning to it; no, as the pressure where the ions travel. A gauge reads the gas at its own location, through its own physics, calibrated for a gas you may not be using. Used the same way every time, it was enough for the documented machines to find their operating windows. The Mullins log is all indicated readings, the numbers the gauge itself displays, from one ion gauge (it ionizes a little of the gas and measures the current) on one port. The note states that with hydrogen flowing the true pressure is about 2.2 times the logged number, and the machine found its beam on the logged numbers (hosted chamber-vacuum note, §11). Four things separate the display from the pressure the ions see: type and range, calibration gas, location, and interference from the magnet or the RF. The ions spend most of each turn inside the dee, where no gauge reads; the difference is small for an open dee, large for one with a solid back and a narrow mouth. Find your window with one gauge used one way; apply its gas correction before comparing with a published number.",
   "short_answer_markdown": "Yes, as a repeatable number for finding a setting and returning to it; no, as the pressure where the ions travel. A gauge reads the gas at its own location, through its own physics, calibrated for a gas you may not be using. Used the same way every time, it was enough for the documented machines to find their operating windows. The Mullins log is all indicated readings, the numbers the gauge itself displays, from one ion gauge (it ionizes a little of the gas and measures the current) on one port. The note states that with hydrogen flowing the true pressure is about 2.2 times the logged number, and the machine found its beam on the logged numbers (hosted chamber-vacuum note, §11). Four things separate the display from the pressure the ions see: type and range, calibration gas, location, and interference from the magnet or the RF. The ions spend most of each turn inside the dee, where no gauge reads; the difference is small for an open dee, large for one with a solid back and a narrow mouth. Find your window with one gauge used one way; apply its gas correction before comparing with a published number.",
   "sources": [
    "dg-1406",
    "dg-1473",
    "dg-1737",
    "dg-1854",
    "dg-1923",
    "dg-449",
    "dg-471",
    "dg-475",
    "dg-476"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/vacuum/#inside-the-dee",
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/calculators/vacuum/",
    "https://cyclotroninfo.com/learn/modeling-tools/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "ground-loops",
    "ground-what-floats",
    "ops-worked-yesterday",
    "vac-pump-rated-better",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-can-i-trust-the-gauge"
  },
  {
   "id": "vac-gas-control",
   "question": "What's the simplest practical way to control the gas flow and the chamber pressure? Needle valve, leak valve or mass-flow controller, and can I leave the setting alone once I find it?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 5,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Control the flow in; the chamber pressure follows from that flow divided by the pumping speed, and it drifts even when the flow does not. They are different quantities even when one instrument can be made to regulate either ( Prechtl Wolf (2020), p. 51-53 ). The documented amateur machines used a needle valve (a fine-taper metering valve), a calibrated leak (a tiny opening of known flow) or a mass-flow controller (it holds the flow you ask for), and all three found beam. They differ in repeatability and drift. The pressure changes whenever the pumping changes: as the walls take up hydrogen and release it, as the pump warms, as a filament ages. A fixed flow is therefore a fixed gas load (gas entering per second) into a drifting system, and the pressure and the beam peak both move during a session. Write down the valve or controller setting and log the chamber pressure on every run, since that is what the beam saw. The hydrogen upstream of whatever valve you choose is a gas-handling problem the safety page covers. Unattended operation with hydrogen flowing is a safety question, not a convenience, and the hydrogen handling, isolation valve and ventilation live on the safety page.",
   "short_answer_markdown": "Control the flow in; the chamber pressure follows from that flow divided by the pumping speed, and it drifts even when the flow does not. They are different quantities even when one instrument can be made to regulate either (dg-1400). The documented amateur machines used a needle valve (a fine-taper metering valve), a calibrated leak (a tiny opening of known flow) or a mass-flow controller (it holds the flow you ask for), and all three found beam. They differ in repeatability and drift. The pressure changes whenever the pumping changes: as the walls take up hydrogen and release it, as the pump warms, as a filament ages. A fixed flow is therefore a fixed gas load (gas entering per second) into a drifting system, and the pressure and the beam peak both move during a session. Write down the valve or controller setting and log the chamber pressure on every run, since that is what the beam saw. The hydrogen upstream of whatever valve you choose is a gas-handling problem the safety page covers. Unattended operation with hydrogen flowing is a safety question, not a convenience, and the hydrogen handling, isolation valve and ventilation live on the safety page.",
   "sources": [
    "dg-1400",
    "dg-1404",
    "dg-1410",
    "dg-1438",
    "dg-1483",
    "dg-1488",
    "dg-1868",
    "dg-2063",
    "dg-388",
    "dg-422",
    "dg-424",
    "dg-425"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/ion-sources/#gas-feed",
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/learn/safety/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-which-ion",
    "ops-the-knobs",
    "ops-worked-yesterday",
    "vac-how-clean",
    "vac-pump-rated-better",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-gas-control"
  },
  {
   "id": "vac-what-materials",
   "question": "What materials can I use inside the chamber? Are ordinary plastics, adhesives, solder, rubber seals and workshop metals acceptable?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 6,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Most of them, judged part by part. Ask which of four jobs the part does, holding something up, insulating, sealing or carrying current; then four properties decide. Is it magnetic? Iron parts in the magnet’s field distort it, so anything there must be non-magnetic, aluminium, copper or brass, unless it is deliberately part of the iron that carries the magnet’s flux. The trap is screws, fittings and clamps of unstated grade; a hand magnet at the bench screens every part ( Dewan (2007), p. 9 ; the vacuum-chamber chapter). Does it give off gas? Rubber and plastic do, and the compounds differ by more than a factor of ten. Will it hold the voltage and the RF? Will it survive the heat where it sits? Judged that way, documented machines used more workshop material than the books suggest. The Mullins chamber is a ring cut from stainless pipe with ten commercial flanges set in epoxy, two Viton (a fluorocarbon rubber) O-rings sealing since 2018, an aluminium dee, a ceramic filament button and silver-plated screws. It has run at 10⁻⁶ to 10⁻⁵ torr for eight years (hosted chamber-vacuum note, §3–§7). Each is a documented use in one application, not a clearance for the material in general.",
   "short_answer_markdown": "Most of them, judged part by part. Ask which of four jobs the part does, holding something up, insulating, sealing or carrying current; then four properties decide. Is it magnetic? Iron parts in the magnet's field distort it, so anything there must be non-magnetic, aluminium, copper or brass, unless it is deliberately part of the iron that carries the magnet's flux. The trap is screws, fittings and clamps of unstated grade; a hand magnet at the bench screens every part (dg-491; the vacuum-chamber chapter). Does it give off gas? Rubber and plastic do, and the compounds differ by more than a factor of ten. Will it hold the voltage and the RF? Will it survive the heat where it sits? Judged that way, documented machines used more workshop material than the books suggest. The Mullins chamber is a ring cut from stainless pipe with ten commercial flanges set in epoxy, two Viton (a fluorocarbon rubber) O-rings sealing since 2018, an aluminium dee, a ceramic filament button and silver-plated screws. It has run at 10⁻⁶ to 10⁻⁵ torr for eight years (hosted chamber-vacuum note, §3–§7). Each is a documented use in one application, not a clearance for the material in general.",
   "sources": [
    "dg-1345",
    "dg-1439",
    "dg-1446",
    "dg-1481",
    "dg-1486",
    "dg-262",
    "dg-298",
    "dg-349",
    "dg-354",
    "dg-355",
    "dg-373",
    "dg-374",
    "dg-437",
    "dg-438",
    "dg-451",
    "dg-454",
    "dg-456",
    "dg-472",
    "dg-478",
    "dg-489",
    "dg-491",
    "dg-550"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/vacuum-chamber/#materials",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#elastomers",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#ports",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#hv-inside",
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "magnet-alignment",
    "rf-two-dees-or-dummy",
    "vac-how-clean",
    "vac-pump-rated-better"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-what-materials"
  },
  {
   "id": "vac-how-clean",
   "question": "How clean do the chamber and the plumbing really need to be? Which of the gloves, the special solvents, the deionized-water rinse and the bake-out do I actually need?",
   "group": "E",
   "group_title": "Vacuum, Materials and Gas",
   "part": 2,
   "order": 7,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Clean of oil, fingerprints and loose debris, dry, and pumped long enough; the rest is for pressures a first machine does not need. Oil, grease and fingerprints give off gas under vacuum for a long time, and so does the water that settles on the walls from the air whenever the chamber is open. Four different activities are called “cleaning”, and the documented amateur machines did the first three and mostly skipped the fourth. Cleaning removes the organic matter; handling, with gloves, keeps fingerprints off afterwards; drying means venting through a desiccant or with dry air so less water comes in ( Cressman (2006), p. 30 , Livingston Boch (1952), p. 44 ). Bake-out, heating the chamber while it is pumped, drives the water out faster; an elastomer-sealed (rubber-sealed) chamber can take at most about 150 °C ( Bertolini (2004), p. 443-446 ). The first three always; the bake when you can; the residue-free (deionized-water) rinse where a surface has to hold voltage, otherwise only when chasing a decade, a factor of ten, you do not need yet. The Mullins machine baked once, in 2017, and never again; after months open to air in 2024, hours of pumping brought its pressure back and beam was recovered the next day (hosted chamber-vacuum note, §13, §14).",
   "short_answer_markdown": "Clean of oil, fingerprints and loose debris, dry, and pumped long enough; the rest is for pressures a first machine does not need. Oil, grease and fingerprints give off gas under vacuum for a long time, and so does the water that settles on the walls from the air whenever the chamber is open. Four different activities are called \"cleaning\", and the documented amateur machines did the first three and mostly skipped the fourth. Cleaning removes the organic matter; handling, with gloves, keeps fingerprints off afterwards; drying means venting through a desiccant or with dry air so less water comes in (dg-1472, dg-474). Bake-out, heating the chamber while it is pumped, drives the water out faster; an elastomer-sealed (rubber-sealed) chamber can take at most about 150 °C (dg-454). The first three always; the bake when you can; the residue-free (deionized-water) rinse where a surface has to hold voltage, otherwise only when chasing a decade, a factor of ten, you do not need yet. The Mullins machine baked once, in 2017, and never again; after months open to air in 2024, hours of pumping brought its pressure back and beam was recovered the next day (hosted chamber-vacuum note, §13, §14).",
   "sources": [
    "dg-1472",
    "dg-253",
    "dg-349",
    "dg-443",
    "dg-444",
    "dg-447",
    "dg-451",
    "dg-452",
    "dg-453",
    "dg-454",
    "dg-456",
    "dg-457",
    "dg-463",
    "dg-466",
    "dg-469",
    "dg-474"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/vacuum/#gas-load",
    "https://cyclotroninfo.com/learn/vacuum/#seals-and-leaks",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#elastomers",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#hv-inside",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "rf-conditioning-or-fault",
    "vac-pump-rated-better",
    "vac-what-materials",
    "vac-what-pump"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/vacuum-materials-and-gas/#vac-how-clean"
  },
  {
   "id": "rf-why-resonate",
   "question": "Why can't I just drive the dee with an RF amplifier, or step the voltage up with a transformer? What are the resonant circuit and the matching network actually doing?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 1,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "You can in principle, but a coil across the dee carries the dee’s current instead, for far less power. The ions circle at a rate the magnet’s field sets, and the dee pushes them at each gap crossing, so its voltage must swing at that rate, the cyclotron frequency. The dee is a capacitor: a plate facing the grounded chamber across a gap, holding charge in proportion to its voltage. Each RF cycle the voltage swings positive and negative, so that charge reverses twice. At kilovolts and megahertz that is tens of amperes in and out for almost no power, since the charge comes back as the voltage falls. A coil across the dee lets the energy swing between them unaided; the amplifier supplies only what each swing loses to resistance. A coil and a capacitor exchange energy at one natural frequency, the resonant frequency. The coil, or a transformer’s winding, is chosen so that frequency is the cyclotron frequency; off it the amplifier carries the current itself. Matching, the last step, arranges the connection so the amplifier delivers its power into the coil and dee. Fix the cyclotron frequency first, then a coil that resonates there with your dee.",
   "short_answer_markdown": "You can in principle, but a coil across the dee carries the dee's current instead, for far less power. The ions circle at a rate the magnet's field sets, and the dee pushes them at each gap crossing, so its voltage must swing at that rate, the cyclotron frequency. The dee is a capacitor: a plate facing the grounded chamber across a gap, holding charge in proportion to its voltage. Each RF cycle the voltage swings positive and negative, so that charge reverses twice. At kilovolts and megahertz that is tens of amperes in and out for almost no power, since the charge comes back as the voltage falls. A coil across the dee lets the energy swing between them unaided; the amplifier supplies only what each swing loses to resistance. A coil and a capacitor exchange energy at one natural frequency, the resonant frequency. The coil, or a transformer's winding, is chosen so that frequency is the cyclotron frequency; off it the amplifier carries the current itself. Matching, the last step, arranges the connection so the amplifier delivers its power into the coil and dee. Fix the cyclotron frequency first, then a coil that resonates there with your dee.",
   "sources": [
    "dg-1390",
    "dg-1857",
    "dg-231",
    "dg-237",
    "dg-240",
    "dg-291",
    "dg-305",
    "dg-323"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/learn/dee-coupling/#loop-or-probe",
    "https://cyclotroninfo.com/learn/dee-coupling/#measuring",
    "https://cyclotroninfo.com/calculators/dee-capacitance-matching/",
    "https://cyclotroninfo.com/design-guide/rf/"
   ],
   "refs": [
    "rf-bench-vs-installed",
    "rf-conditioning-or-fault",
    "rf-loop-or-tap",
    "rf-watts-to-volts",
    "rf-what-is-q",
    "rf-which-resonance"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-why-resonate"
  },
  {
   "id": "rf-which-resonance",
   "question": "I keep seeing \"cyclotron resonance\", \"RF resonance\" and \"impedance matching\". Are these the same thing? What has to match what?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 2,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Three different things. A first beam needs the first two; the third delivers the amplifier’s power efficiently and keeps the amplifier within its limits. Cyclotron resonance is between the ions and the RF: the dee voltage has to reverse in step with the ions’ circling, which the magnet’s field sets. RF resonance is inside your electronics. The dee is a metal electrode close to grounded metal, so it is one plate of a capacitor, two conductors storing charge. That capacitance and a coil form a tank circuit, whose stored energy swings between coil and capacitor at one natural frequency, and only at that frequency does a small amplifier produce kilovolts ( Why can't I just drive the dee with an RF amplifier, or step the voltage up with a transformer? What are the resonant circuit and the matching network actually doing? ). Matching is between the amplifier and the tank. The amplifier is built to drive 50 Ω and the tank at resonance looks like a different resistance. The coupling, the loop or capacitor connecting them, transforms one into the other so that power goes into the tank rather than back toward the amplifier. Matching is not a condition for acceleration. Tune the RF resonance to the cyclotron frequency, the ions’ circling rate, and make the match at the RF resonance; the two interact, so there is no fixed order.",
   "short_answer_markdown": "Three different things. A first beam needs the first two; the third delivers the amplifier's power efficiently and keeps the amplifier within its limits. *Cyclotron resonance* is between the ions and the RF: the dee voltage has to reverse in step with the ions' circling, which the magnet's field sets. *RF resonance* is inside your electronics. The dee is a metal electrode close to grounded metal, so it is one plate of a capacitor, two conductors storing charge. That capacitance and a coil form a tank circuit, whose stored energy swings between coil and capacitor at one natural frequency, and only at that frequency does a small amplifier produce kilovolts (rf-why-resonate). *Matching* is between the amplifier and the tank. The amplifier is built to drive 50 Ω and the tank at resonance looks like a different resistance. The coupling, the loop or capacitor connecting them, transforms one into the other so that power goes into the tank rather than back toward the amplifier. Matching is not a condition for acceleration. Tune the RF resonance to the cyclotron frequency, the ions' circling rate, and make the match at the RF resonance; the two interact, so there is no fixed order.",
   "sources": [
    "dg-1392",
    "dg-1424",
    "dg-1809",
    "dg-236",
    "dg-238",
    "dg-302",
    "dg-304",
    "dg-305",
    "dg-306"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#resonance-condition",
    "https://cyclotroninfo.com/learn/exercises/find-the-resonance/",
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/learn/dee-coupling/#practice",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-which-ion",
    "goal-how-it-fits-together",
    "ops-finding-first-beam",
    "rf-bench-vs-installed",
    "rf-loop-or-tap",
    "rf-what-is-q",
    "rf-why-resonate"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-which-resonance"
  },
  {
   "id": "rf-two-dees-or-dummy",
   "question": "Do I need two dees? What does a dummy dee do? Can the chamber lid serve as the RF return, or is that a different issue?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 3,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "One dee driven against ground is what every amateur machine whose RF layout is documented ran; the “dummy dee” is a shaped piece of that ground; and the lids and chamber are already the RF return. The textbook’s two hollow D-shaped electrodes, the ions circling inside them, are driven in opposite phase, one positive while the other is negative, so the gap voltage swings twice as far as either dee does. Driving one dee against a grounded electrode halves the feedthroughs, the insulated RF connections through the chamber wall, and makes the chamber the return electrode, the path the RF current takes back to the coil ( Dewan (2007), p. 13 , Wouters (1949), p. 8 ). The Mullins machine has no dummy dee; its 2018 build log says the grounded lids are close enough to serve, and nothing later changed that (hosted rf-dee-source note, §2.1). A dummy dee, a grounded edge shaped to match the dee’s, evens out the electric field the first turns cross and adds capacitance that lowers the tuning. The return current runs through lids, wall and flanges in series; their resistance lowers the volts each watt buys ( What is this Q people talk about? How do I measure it, how good does it need to be, and is higher always better? ). For a first beam, one dee against ground with a clean, clamped return is enough; the dummy dee can follow.",
   "short_answer_markdown": "One dee driven against ground is what every amateur machine whose RF layout is documented ran; the \"dummy dee\" is a shaped piece of that ground; and the lids and chamber are already the RF return. The textbook's two hollow D-shaped electrodes, the ions circling inside them, are driven in opposite phase, one positive while the other is negative, so the gap voltage swings twice as far as either dee does. Driving one dee against a grounded electrode halves the feedthroughs, the insulated RF connections through the chamber wall, and makes the chamber the return electrode, the path the RF current takes back to the coil (dg-258, dg-278). The Mullins machine has no dummy dee; its 2018 build log says the grounded lids are close enough to serve, and nothing later changed that (hosted rf-dee-source note, §2.1). A dummy dee, a grounded edge shaped to match the dee's, evens out the electric field the first turns cross and adds capacitance that lowers the tuning. The return current runs through lids, wall and flanges in series; their resistance lowers the volts each watt buys (rf-what-is-q). For a first beam, one dee against ground with a clean, clamped return is enough; the dummy dee can follow.",
   "sources": [
    "dg-1003",
    "dg-1352",
    "dg-1395",
    "dg-1440",
    "dg-1743",
    "dg-1852",
    "dg-240",
    "dg-258",
    "dg-278",
    "dg-283",
    "dg-284",
    "dg-311",
    "dg-317",
    "dg-322",
    "dg-570",
    "dg-667",
    "dg-674"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/learn/dee-coupling/#discharges",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#lids",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#hv-inside",
    "https://cyclotroninfo.com/library/amateur/#mullins",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "ground-what-floats",
    "magnet-alignment",
    "rf-bench-vs-installed",
    "rf-what-is-q",
    "rf-why-resonate",
    "source-where-does-it-go"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-two-dees-or-dummy"
  },
  {
   "id": "rf-is-the-chamber-a-cavity",
   "question": "Does my vacuum chamber itself have to be a resonant cavity? What is a resonant cavity, and how does it relate to the dee's resonant circuit?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 4,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "No. On a small machine the resonant circuit is a coil and the dee’s capacitance, the charge the dee stores per volt against the grounded metal around it; the chamber is the other plate of that capacitor and the path the current returns along. The word “cavity” describes the resonant structure of a large or high-frequency machine. When the structure is a sizeable fraction of an RF wavelength, the distance the wave travels in one cycle, the voltage differs along the dee itself, and the designers treat the whole structure as the resonant element, its frequency set by its length ( Livingston Howard (eds.), p. 17 , Livingston Blewett (1962) ). At a tabletop machine’s 5 to 20 MHz a wavelength is 15 to 60 metres. A chamber a few tens of centimetres across is too small to resonate at those frequencies. It supplies capacitance instead: Wouters’ 1949 memo says the dee-to-ground capacity “appears as the major portion of the capacitance in the tank circuit”, the coil and dee together ( Wouters (1949), p. 8 ), and the documented amateur machines resonated coils outside the vacuum against that capacitance. Measure your dee’s capacitance with everything in place, and choose the coil to resonate with it at the ions’ circling frequency (the matching calculator).",
   "short_answer_markdown": "No. On a small machine the resonant circuit is a coil and the dee's capacitance, the charge the dee stores per volt against the grounded metal around it; the chamber is the other plate of that capacitor and the path the current returns along. The word \"cavity\" describes the resonant structure of a large or high-frequency machine. When the structure is a sizeable fraction of an RF wavelength, the distance the wave travels in one cycle, the voltage differs along the dee itself, and the designers treat the whole structure as the resonant element, its frequency set by its length (dg-323, dg-250). At a tabletop machine's 5 to 20 MHz a wavelength is 15 to 60 metres. A chamber a few tens of centimetres across is too small to resonate at those frequencies. It supplies capacitance instead: Wouters' 1949 memo says the dee-to-ground capacity \"appears as the major portion of the capacitance in the tank circuit\", the coil and dee together (dg-279), and the documented amateur machines resonated coils outside the vacuum against that capacitance. Measure your dee's capacitance with everything in place, and choose the coil to resonate with it at the ions' circling frequency (the matching calculator).",
   "sources": [
    "dg-1003",
    "dg-1097",
    "dg-1128",
    "dg-1130",
    "dg-1131",
    "dg-1369",
    "dg-1390",
    "dg-1392",
    "dg-1857",
    "dg-250",
    "dg-279",
    "dg-306",
    "dg-323",
    "dg-664",
    "dg-999"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/calculators/dee-capacitance-matching/",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#hv-inside",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "rf-bench-vs-installed",
    "rf-two-dees-or-dummy",
    "rf-which-resonance",
    "rf-why-resonate"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-is-the-chamber-a-cavity"
  },
  {
   "id": "rf-what-is-q",
   "question": "What is this \"Q\" people talk about? How do I measure it, how good does it need to be, and is higher always better?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 5,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Q is how many cycles the tank, the coil and dee together, keeps ringing after you stop pushing it, give or take a factor of six. The tank stores energy, swinging it between dee and coil, and loses a little each cycle to resistance; Q is the stored energy divided by the energy lost per cycle, times two pi. High Q means little loss, so a small amplifier can keep a large swing going; like for like, the dee voltage for a given power rises as the square root of Q ( Why can't I just drive the dee with an RF amplifier, or step the voltage up with a transformer? What are the resonant circuit and the matching network actually doing? ). Measure it by sweeping the frequency across the resonance; centre frequency divided by peak width is Q. Two Qs exist, the tank’s own and the lower one read with the amplifier connected, so every quoted Q carries its configuration. The documented values span two orders of magnitude and both ends made beam: Houghton’s antenna-tuner chain measured 16 ( Haas (2009), p. 64 ), the Rutgers 12-inch’s copper tank 460 through its coupling at match ( Koeth (2005), p. 6-8 ). Higher is not always better: the peak narrows with it, and drift moves you off it sooner. Find beam with the Q you have if the amplifier supplies the watts; raise it later for more volts per watt.",
   "short_answer_markdown": "Q is how many cycles the tank, the coil and dee together, keeps ringing after you stop pushing it, give or take a factor of six. The tank stores energy, swinging it between dee and coil, and loses a little each cycle to resistance; Q is the stored energy divided by the energy lost per cycle, times two pi. High Q means little loss, so a small amplifier can keep a large swing going; like for like, the dee voltage for a given power rises as the square root of Q (rf-why-resonate). Measure it by sweeping the frequency across the resonance; centre frequency divided by peak width is Q. Two Qs exist, the tank's own and the lower one read with the amplifier connected, so every quoted Q carries its configuration. The documented values span two orders of magnitude and both ends made beam: Houghton's antenna-tuner chain measured 16 (dg-305), the Rutgers 12-inch's copper tank 460 through its coupling at match (dg-237). Higher is not always better: the peak narrows with it, and drift moves you off it sooner. Find beam with the Q you have if the amplifier supplies the watts; raise it later for more volts per watt.",
   "sources": [
    "dg-1859",
    "dg-222",
    "dg-224",
    "dg-227",
    "dg-228",
    "dg-236",
    "dg-237",
    "dg-240",
    "dg-275",
    "dg-305",
    "dg-306",
    "dg-313",
    "dg-317",
    "dg-332",
    "dg-674"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/learn/dee-coupling/#measuring",
    "https://cyclotroninfo.com/calculators/dee-capacitance-matching/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "rf-bench-vs-installed",
    "rf-conditioning-or-fault",
    "rf-loop-or-tap",
    "rf-watts-to-volts",
    "rf-why-resonate"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-what-is-q"
  },
  {
   "id": "rf-loop-or-tap",
   "question": "Should I use inductive or capacitive coupling to drive the dee? Which is easier to build, adjust and troubleshoot?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 6,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Either, and the one with a precedent you can copy is the easier one. A coupling loop is a turn or two of wire close to the tank coil, linked to it by the coil’s magnetic field as in a transformer. A capacitive tap is a wire into the dee side of the tank through a series capacitor, or a connection part-way along the coil. Both are impedance transformers. The tank, the coil and dee together, looks like a large resistance at resonance and the amplifier is built to drive 50 Ω; either arrangement makes the one look like the other, so on a given tank they are equivalent once matched. On the Rutgers resonator every loop spacing and tap setting that presented 50 Ω gave the same dee voltage for the same power, so the group chose on mechanical grounds ( Koeth (2005), p. 7 ). The loop is adjusted by moving it and keeps any DC voltage on the dee off the amplifier; the tap is adjusted by a variable capacitor, which sits in the high-voltage path. Choose whichever you can match reliably with your amplifier, and protect the amplifier regardless, since the power a mismatched tank refuses comes back into it.",
   "short_answer_markdown": "Either, and the one with a precedent you can copy is the easier one. A coupling loop is a turn or two of wire close to the tank coil, linked to it by the coil's magnetic field as in a transformer. A capacitive tap is a wire into the dee side of the tank through a series capacitor, or a connection part-way along the coil. Both are impedance transformers. The tank, the coil and dee together, looks like a large resistance at resonance and the amplifier is built to drive 50 Ω; either arrangement makes the one look like the other, so on a given tank they are equivalent once matched. On the Rutgers resonator every loop spacing and tap setting that presented 50 Ω gave the same dee voltage for the same power, so the group chose on mechanical grounds (dg-238). The loop is adjusted by moving it and keeps any DC voltage on the dee off the amplifier; the tap is adjusted by a variable capacitor, which sits in the high-voltage path. Choose whichever you can match reliably with your amplifier, and protect the amplifier regardless, since the power a mismatched tank refuses comes back into it.",
   "sources": [
    "dg-1392",
    "dg-1857",
    "dg-235",
    "dg-236",
    "dg-238",
    "dg-285",
    "dg-305",
    "dg-904"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#loop-or-probe",
    "https://cyclotroninfo.com/learn/dee-coupling/#practice",
    "https://cyclotroninfo.com/calculators/dee-capacitance-matching/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "rf-bench-vs-installed",
    "rf-watts-to-volts",
    "rf-what-is-q",
    "rf-which-resonance",
    "rf-why-resonate"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-loop-or-tap"
  },
  {
   "id": "rf-watts-to-volts",
   "question": "How much RF power do I need, and how does that become voltage on the dee? Can I read the voltage off the power meter, or do I have to measure it?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 7,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Tens of watts have made first beams on several documented machines, and the power meter cannot tell you the voltage. Houghton’s commissioning run made a measurable peak at 15 W forward, forward power being what the amplifier sends toward the tank, the coil and dee together ( Haas (2009), p. 69 ); the Rutgers 12-inch, chasing kilovolts, runs 500 W ( Koeth (2007), p. 3 ). Nearly all the power reaching the tank is lost as heat in the resistance of the coil, joints and return path. That loss goes as the square of the voltage, so at a fixed match, the setting that lets the amplifier’s power into the tank, the dee voltage rises as the square root of forward power (the matching calculator). Doubling the voltage costs four times the power; the constant is your machine’s losses, so the same watts give different volts on different machines. The meter reads forward power minus reflected power, the part that comes back toward the amplifier: the power accepted where the meter sits. A low reflected reading says the amplifier is matched and does not say where the power went, since a lossy tuner matches perfectly and delivers few volts ( Haas (2009), p. 64 ). To know the voltage, measure it, with a probe or a calibrated pickup, or infer it from the beam.",
   "short_answer_markdown": "Tens of watts have made first beams on several documented machines, and the power meter cannot tell you the voltage. Houghton's commissioning run made a measurable peak at 15 W forward, forward power being what the amplifier sends toward the tank, the coil and dee together (dg-308); the Rutgers 12-inch, chasing kilovolts, runs 500 W (dg-1647). Nearly all the power reaching the tank is lost as heat in the resistance of the coil, joints and return path. That loss goes as the square of the voltage, so at a fixed match, the setting that lets the amplifier's power into the tank, the dee voltage rises as the square root of forward power (the matching calculator). Doubling the voltage costs four times the power; the constant is your machine's losses, so the same watts give different volts on different machines. The meter reads forward power minus reflected power, the part that comes back toward the amplifier: the power accepted where the meter sits. A low reflected reading says the amplifier is matched and does not say where the power went, since a lossy tuner matches perfectly and delivers few volts (dg-305). To know the voltage, measure it, with a probe or a calibrated pickup, or infer it from the beam.",
   "sources": [
    "dg-1368",
    "dg-1391",
    "dg-1647",
    "dg-1860",
    "dg-1861",
    "dg-2052",
    "dg-2053",
    "dg-231",
    "dg-233",
    "dg-239",
    "dg-288",
    "dg-291",
    "dg-294",
    "dg-305",
    "dg-306",
    "dg-307",
    "dg-308",
    "dg-313"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#resonant-load",
    "https://cyclotroninfo.com/learn/dee-coupling/#measuring",
    "https://cyclotroninfo.com/calculators/dee-capacitance-matching/",
    "https://cyclotroninfo.com/learn/safety/#hazard-xrays",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "rf-conditioning-or-fault",
    "rf-sub-kilovolt-dees",
    "rf-what-is-q",
    "rf-why-resonate",
    "safety-before-beam",
    "source-how-many-turns"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-watts-to-volts"
  },
  {
   "id": "rf-sub-kilovolt-dees",
   "question": "The literature says tens of kilovolts on the dee. Some amateur machines ran under a kilovolt. How is that possible, and what did it cost them?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 8,
   "tags": [
    "build"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "The dee voltage does not set the energy, and the low voltage cost them turns. The energy an ion reaches is set by the magnetic field and the radius, since an ion circles on a radius that grows with its energy; the dee voltage sets how many pushes it takes to get there ( Prechtl Wolf (2020), p. 41-42 ; How many turns should a proton make before it reaches my target energy? Fewer turns sounds easier, but isn't the whole point of a cyclotron to accelerate over lots of turns? ). The amateur first beams ran an order of magnitude below the literature’s tens of kilovolts: the Mullins dee is “~800 V nominal, uncalibrated” (hosted rf-dee-source note, §5). An ion crosses the gap twice a turn and gains the gap voltage at each crossing, so at a kilovolt a 150 keV proton (150 kilovolts of pushes) makes about a hundred turns rather than ten. Every extra turn is more path through the gas and more chance to fall out of step with the RF. The site treats any gap above about 15 kV as an operating X-ray source, and an RF system that will hold kilovolts is surveyed from its first power-up, before any voltage is raised ( What safety checks, expert review and local permissions do I need before I start testing the subsystems? ; Can the machine make radiation, or hurt me some other way, before I've even detected a beam? ). The low voltage is enough for a first beam; more buys fewer turns, at the square root of the power ( How much RF power do I need, and how does that become voltage on the dee? Can I read the voltage off the power meter, or do I have to measure it? ; the path-length calculator).",
   "short_answer_markdown": "The dee voltage does not set the energy, and the low voltage cost them turns. The energy an ion reaches is set by the magnetic field and the radius, since an ion circles on a radius that grows with its energy; the dee voltage sets how many pushes it takes to get there (dg-1386; source-how-many-turns). The amateur first beams ran an order of magnitude below the literature's tens of kilovolts: the Mullins dee is \"~800 V nominal, uncalibrated\" (hosted rf-dee-source note, §5). An ion crosses the gap twice a turn and gains the gap voltage at each crossing, so at a kilovolt a 150 keV proton (150 kilovolts of pushes) makes about a hundred turns rather than ten. Every extra turn is more path through the gas and more chance to fall out of step with the RF. The site treats any gap above about 15 kV as an operating X-ray source, and an RF system that will hold kilovolts is surveyed from its first power-up, before any voltage is raised (safety-before-testing; safety-before-beam). The low voltage is enough for a first beam; more buys fewer turns, at the square root of the power (rf-watts-to-volts; the path-length calculator).",
   "sources": [
    "dg-026",
    "dg-1280",
    "dg-1364",
    "dg-1386",
    "dg-1387",
    "dg-1388",
    "dg-1389",
    "dg-1399",
    "dg-1423",
    "dg-1647",
    "dg-1735",
    "dg-1809",
    "dg-181",
    "dg-1861",
    "dg-231",
    "dg-246",
    "dg-270",
    "dg-288",
    "dg-289",
    "dg-311",
    "dg-582",
    "dg-680",
    "dg-698"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/how-a-cyclotron-works/#dees-and-gap",
    "https://cyclotroninfo.com/learn/exercises/more-voltage-same-energy/",
    "https://cyclotroninfo.com/calculators/path-length/",
    "https://cyclotroninfo.com/learn/dee-coupling/#discharges",
    "https://cyclotroninfo.com/learn/safety/#hazard-xrays",
    "https://cyclotroninfo.com/library/amateur/#mullins",
    "https://cyclotroninfo.com/builds/"
   ],
   "refs": [
    "rf-conditioning-or-fault",
    "rf-watts-to-volts",
    "safety-before-beam",
    "source-how-many-turns",
    "source-timing-with-rf",
    "source-where-does-it-go",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-sub-kilovolt-dees"
  },
  {
   "id": "rf-bench-vs-installed",
   "question": "My RF circuit tunes up nicely on the bench. Why does it change when I close the chamber, pump down, turn on the source, or raise the power?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 9,
   "tags": [
    "build"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Because the bench measured a different circuit. The tank, the coil and the dee together, resonates at a frequency set by the coil and by the capacitance the dee sees, its charge per volt against nearby grounded metal. The match, the setting that lets the amplifier’s power in, is made at that frequency, so a few percent of capacitance moves both. Geometry. Closing the chamber brings grounded metal closer to the dee and pumping down bows the lids toward it; capacitance rises as plates approach, so the resonance drops. Electrical loading. Gas, a running source or, at higher power, a sparking surface puts a conducting medium near the dee; a discharge loads the tank and holds the voltage down. Temperature. RF current heats coil and dee, the metal expands, the resonance walks. Measurement conditions. The reflected-power meter reads the match at its own place in the chain. The Mullins chain was developed on the bench against a “simulated chamber” capacitor of 387 pF and met a terminal capacitance of 128 pF (hosted rf-dee-source note §2.3, §2.4, §4.1). Retune after assembly, pump-down and warm-up, and log which way and how far the resonance moved; direction and size point at the cause.",
   "short_answer_markdown": "Because the bench measured a different circuit. The tank, the coil and the dee together, resonates at a frequency set by the coil and by the capacitance the dee sees, its charge per volt against nearby grounded metal. The match, the setting that lets the amplifier's power in, is made at that frequency, so a few percent of capacitance moves both. *Geometry.* Closing the chamber brings grounded metal closer to the dee and pumping down bows the lids toward it; capacitance rises as plates approach, so the resonance drops. *Electrical loading.* Gas, a running source or, at higher power, a sparking surface puts a conducting medium near the dee; a discharge loads the tank and holds the voltage down. *Temperature.* RF current heats coil and dee, the metal expands, the resonance walks. *Measurement conditions.* The reflected-power meter reads the match at its own place in the chain. The Mullins chain was developed on the bench against a \"simulated chamber\" capacitor of 387 pF and met a terminal capacitance of 128 pF (hosted rf-dee-source note §2.3, §2.4, §4.1). Retune after assembly, pump-down and warm-up, and log which way and how far the resonance moved; direction and size point at the cause.",
   "sources": [
    "dg-1003",
    "dg-1858",
    "dg-1862",
    "dg-187",
    "dg-251",
    "dg-252",
    "dg-293",
    "dg-306",
    "dg-680",
    "dg-725",
    "dg-999"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#architecture",
    "https://cyclotroninfo.com/learn/dee-coupling/#practice",
    "https://cyclotroninfo.com/learn/vacuum-chamber/#lids",
    "https://cyclotroninfo.com/calculators/lid-deflection/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "magnet-alignment",
    "ops-worked-yesterday",
    "rf-conditioning-or-fault",
    "rf-loop-or-tap",
    "rf-what-is-q",
    "rf-which-resonance"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-bench-vs-installed"
  },
  {
   "id": "rf-conditioning-or-fault",
   "question": "People talk about RF \"conditioning\". What does that mean, and how do I tell conditioning from a fault that means I should stop?",
   "group": "F",
   "group_title": "The Dee and Its RF",
   "part": 2,
   "order": 10,
   "tags": [
    "build",
    "run"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "RF conditioning is a controlled process that raises how much voltage a vacuum RF structure holds. Electrodes recently vented, opened to air, carry contaminants, gas and microscopic points that spark well below what the gap will eventually hold; some of those effects diminish under controlled operation. An unexpected discharge is not automatically conditioning. The distinction cannot be made simply by watching whether the sparking becomes less frequent: a fault can also appear to improve for a while, and a falling spark rate does not establish that the equipment is safe or undamaged. Wouters’ 1949 small-machine procedure sets limits. Apply RF in short bursts at low power, raise the power and the duration gradually, never leave the RF on through a discharge, and treat the job as done when the vacuum stays below about 10⁻⁴ torr with a couple of kilovolts on steadily ( Wouters (1949) ). Before the first burst come the prerequisites, which do not relax. The chamber assembled clean, with no grease or fingerprints on the high-field surfaces and nothing abrasive used on them ( Livingston Blewett (1962), p. 189 ; How clean do the chamber and the plumbing really need to be? Which of the gloves, the special solvents, the deionized-water rinse and the bake-out do I actually need? ). The pressure in the 10⁻⁵ decade. The amplifier protected against the mismatch a sparking load presents, by reflected-power trips or a final that survives an open load. Hands off the live system. And the X-ray survey instrument on from the RF system’s first power-up where the protection plan requires one, because a sparking vacuum gap at tens of kilovolts is an X-ray source and the plan covers any subsystem that can make one ( What safety checks, expert review and local permissions do I need before I start testing the subsystems? ; Can the machine make radiation, or hurt me some other way, before I've even detected a beam? ).",
   "short_answer_markdown": "RF conditioning is a controlled process that raises how much voltage a vacuum RF structure holds. Electrodes recently vented, opened to air, carry contaminants, gas and microscopic points that spark well below what the gap will eventually hold; some of those effects diminish under controlled operation. An unexpected discharge is not automatically conditioning. The distinction cannot be made simply by watching whether the sparking becomes less frequent: a fault can also appear to improve for a while, and a falling spark rate does not establish that the equipment is safe or undamaged. Wouters' 1949 small-machine procedure sets limits. Apply RF in short bursts at low power, raise the power and the duration gradually, never leave the RF on through a discharge, and treat the job as done when the vacuum stays below about 10⁻⁴ torr with a couple of kilovolts on steadily (dg-277). Before the first burst come the prerequisites, which do not relax. The chamber assembled clean, with no grease or fingerprints on the high-field surfaces and nothing abrasive used on them (dg-253; vac-how-clean). The pressure in the 10⁻⁵ decade. The amplifier protected against the mismatch a sparking load presents, by reflected-power trips or a final that survives an open load. Hands off the live system. And the X-ray survey instrument on from the RF system's first power-up where the protection plan requires one, because a sparking vacuum gap at tens of kilovolts is an X-ray source and the plan covers any subsystem that can make one (safety-before-testing; safety-before-beam).",
   "sources": [
    "dg-1274",
    "dg-1280",
    "dg-1364",
    "dg-1365",
    "dg-1366",
    "dg-1371",
    "dg-1987",
    "dg-2013",
    "dg-2020",
    "dg-2036",
    "dg-222",
    "dg-251",
    "dg-253",
    "dg-276",
    "dg-277",
    "dg-285",
    "dg-315",
    "dg-351",
    "dg-667",
    "dg-680",
    "dg-682",
    "dg-783",
    "dg-935"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/dee-coupling/#discharges",
    "https://cyclotroninfo.com/learn/safety/#hazard-electrical",
    "https://cyclotroninfo.com/learn/safety/#hazard-xrays",
    "https://cyclotroninfo.com/learn/vacuum/#voltage-holding",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "rf-bench-vs-installed",
    "rf-sub-kilovolt-dees",
    "rf-watts-to-volts",
    "safety-before-beam",
    "safety-before-testing",
    "vac-how-clean"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/the-dee-and-its-rf/#rf-conditioning-or-fault"
  },
  {
   "id": "ops-the-knobs",
   "question": "What are all the knobs, and what does each one actually change? Which are operating controls, which are setup adjustments, and which can I leave alone?",
   "group": "G",
   "group_title": "Finding and Keeping a Beam",
   "part": 3,
   "order": 1,
   "tags": [
    "run"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "A small cyclotron has several adjustments, but you do not normally need to keep changing all of them. Magnet current commands the magnetic field, which sets how fast the ions circle and so where the resonance falls. RF frequency commands where the tank (the RF resonant circuit, of which the dee is the capacitor) is driven; the dee voltage follows only if the tank is tuned there. RF drive level commands the amplifier’s input; the dee voltage follows as the square root of the delivered power. Filament current commands the heater; what changes is emission, the electrons the hot wire gives off. Filament bias commands the voltage between the filament and its surroundings, and so the energy of the electrons that make the ions. Gas flow commands the inlet; the chamber pressure and the source’s output follow. Collector bias and position are primarily measurement adjustments: a voltage on the plate that keeps the reading meaning ion current, and the radius you are reading. Which of those need frequent attention varies by machine and by stage. During a run on the documented machines the magnet, the gas, and sometimes the filament and the RF drive were the ones moved.",
   "short_answer_markdown": "A small cyclotron has several adjustments, but you do not normally need to keep changing all of them. *Magnet current* commands the magnetic field, which sets how fast the ions circle and so where the resonance falls. *RF frequency* commands where the tank (the RF resonant circuit, of which the dee is the capacitor) is driven; the dee voltage follows only if the tank is tuned there. *RF drive level* commands the amplifier's input; the dee voltage follows as the square root of the delivered power. *Filament current* commands the heater; what changes is emission, the electrons the hot wire gives off. *Filament bias* commands the voltage between the filament and its surroundings, and so the energy of the electrons that make the ions. *Gas flow* commands the inlet; the chamber pressure and the source's output follow. *Collector bias and position* are primarily measurement adjustments: a voltage on the plate that keeps the reading meaning ion current, and the radius you are reading. Which of those need frequent attention varies by machine and by stage. During a run on the documented machines the magnet, the gas, and sometimes the filament and the RF drive were the ones moved.",
   "sources": [
    "dg-009",
    "dg-027",
    "dg-1358",
    "dg-1582",
    "dg-1737",
    "dg-2058",
    "dg-2070",
    "dg-2071",
    "dg-243",
    "dg-281",
    "dg-293",
    "dg-377",
    "dg-402",
    "dg-408",
    "dg-428",
    "dg-804"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#skills",
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/learn/dee-coupling/#practice",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "magnet-what-to-measure",
    "ops-finding-first-beam",
    "ops-worked-yesterday",
    "ops-would-my-detector-see-it",
    "rf-watts-to-volts",
    "rf-which-resonance",
    "source-is-it-making-ions",
    "vac-gas-control"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/finding-and-keeping-a-beam/#ops-the-knobs"
  },
  {
   "id": "ops-what-can-i-test",
   "question": "What can I test before the whole cyclotron works? How do I check each subsystem without already having a beam?",
   "group": "G",
   "group_title": "Finding and Keeping a Beam",
   "part": 3,
   "order": 2,
   "tags": [
    "run"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Almost everything, and the documented machines did. A cyclotron that will not make beam has five subsystems, any of which may be fine. Testing them apart separates the failing one from the working ones. The vacuum system is tested empty: the pressure is what the pump alone reaches plus what each part adds, so read it with the pump blanked off, then with each part added in turn. The magnetic field is read against the coil current: the RF frequency fixes the field it must reach, and the readings show whether the supply reaches it. The filament’s emission, the electrons the hot wire gives off, is measured with no RF by a collector held positive to draw them. The dee’s resonant circuit is tuned on the bench against a capacitor standing in for the dee, then re-found installed, since the real dee and lids differ in capacitance. The detector and meter are read with the RF on and the gas off, which starves the source, or the field far from resonance, which leaves the ions out of step with the RF; a beam reading must stand above that floor. None of these shows that the assembled machine will accelerate.",
   "short_answer_markdown": "Almost everything, and the documented machines did. A cyclotron that will not make beam has five subsystems, any of which may be fine. Testing them apart separates the failing one from the working ones. The vacuum system is tested empty: the pressure is what the pump alone reaches plus what each part adds, so read it with the pump blanked off, then with each part added in turn. The magnetic field is read against the coil current: the RF frequency fixes the field it must reach, and the readings show whether the supply reaches it. The filament's emission, the electrons the hot wire gives off, is measured with no RF by a collector held positive to draw them. The dee's resonant circuit is tuned on the bench against a capacitor standing in for the dee, then re-found installed, since the real dee and lids differ in capacitance. The detector and meter are read with the RF on and the gas off, which starves the source, or the field far from resonance, which leaves the ions out of step with the RF; a beam reading must stand above that floor. None of these shows that the assembled machine will accelerate.",
   "sources": [
    "dg-090",
    "dg-1100",
    "dg-1280",
    "dg-1359",
    "dg-1582",
    "dg-1679",
    "dg-1814",
    "dg-367",
    "dg-807",
    "dg-818",
    "dg-952",
    "dg-999"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#timeline",
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/learn/ion-sources/#filament-sources",
    "https://cyclotroninfo.com/learn/dee-coupling/#discharges",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "magnet-what-to-measure",
    "ops-finding-first-beam",
    "ops-would-my-detector-see-it",
    "rf-bench-vs-installed",
    "rf-watts-to-volts",
    "safety-before-testing",
    "source-is-it-making-ions",
    "vac-pump-rated-better"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/finding-and-keeping-a-beam/#ops-what-can-i-test"
  },
  {
   "id": "ops-would-my-detector-see-it",
   "question": "How do I know my beam-current detector could detect the beam I'm looking for, before I use it to troubleshoot the cyclotron?",
   "group": "G",
   "group_title": "Finding and Keeping a Beam",
   "part": 3,
   "order": 3,
   "tags": [
    "run"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "By testing the chain against everything that is not beam, before using it to look for beam. A first beam is picoamperes to nanoamperes on a plate a few centimetres from a dee carrying kilovolts of RF. Every nanoampere of RF rectification (RF picked up by the lead and turned into steady current), leakage across an insulator, or secondary electrons (electrons the ions knock out of the plate, each one that leaves counting as extra positive current) is large against it (the beam-measurement chapter). All three are present with no beam, so run everything except what makes beam, RF at operating power, gas off or magnet far from resonance, and read the meter. Neither condition guarantees that no charged particle reaches the plate; the reading establishes what the chain shows when accelerated beam should be absent, and a peak has to stand above it. The Mullins machine’s open plate and clip-lead chain gave no absolute current better than an order-of-unity estimate (a factor of a few) and found beam anyway (hosted instrumentation note, §3, §5). For a first beam the detector has to answer “is there a peak that comes and goes with the field”; “how many nanoamperes” comes later.",
   "short_answer_markdown": "By testing the chain against everything that is not beam, before using it to look for beam. A first beam is picoamperes to nanoamperes on a plate a few centimetres from a dee carrying kilovolts of RF. Every nanoampere of RF rectification (RF picked up by the lead and turned into steady current), leakage across an insulator, or secondary electrons (electrons the ions knock out of the plate, each one that leaves counting as extra positive current) is large against it (the beam-measurement chapter). All three are present with no beam, so run everything except what makes beam, RF at operating power, gas off or magnet far from resonance, and read the meter. Neither condition guarantees that no charged particle reaches the plate; the reading establishes what the chain shows when accelerated beam should be absent, and a peak has to stand above it. The Mullins machine's open plate and clip-lead chain gave no absolute current better than an order-of-unity estimate (a factor of a few) and found beam anyway (hosted instrumentation note, §3, §5). For a first beam the detector has to answer \"is there a peak that comes and goes with the field\"; \"how many nanoamperes\" comes later.",
   "sources": [
    "dg-1736",
    "dg-241",
    "dg-494",
    "dg-516",
    "dg-517",
    "dg-523",
    "dg-524",
    "dg-736"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#how-small",
    "https://cyclotroninfo.com/learn/beam-measurement/#faraday-cup",
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-documenting-the-claim",
    "beam-is-it-working",
    "ground-loops",
    "ground-what-floats",
    "ops-beam-near-center-not-farther-out",
    "ops-finding-first-beam"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/finding-and-keeping-a-beam/#ops-would-my-detector-see-it"
  },
  {
   "id": "ops-finding-first-beam",
   "question": "How do I find a first beam when all these settings affect one another? Does everything have to be perfect at once before anything happens?",
   "group": "G",
   "group_title": "Finding and Keeping a Beam",
   "part": 3,
   "order": 4,
   "tags": [
    "run"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "No. Most of the settings are established one at a time with no beam. Two of them, the magnetic field and the source (the filament and gas that make the ions), are explored together. The field sets how fast the ions circle; the RF voltage swings at a fixed rate. Only when the two rates match does an ion reach the gap at the same point of every RF swing, in step with it. Then it gets a push at every crossing. That match is the resonance, and it falls at one field value. So the search is a sweep. With the RF fixed, run the magnet current slowly through the expected field and watch the collector. A peak that comes back at the same field on every pass is the signal. The record of first beams is a record of machines that were far from perfect. The Mullins machine found its first resonances in 2018 with a dee of uncalibrated voltage, an unshielded plate and a filament resting loose on the chamber floor (hosted instrumentation note, §7; the build log’s “It works!” page). If no peak appears, reconsider the other conditions, the source above all, and repeat.",
   "short_answer_markdown": "No. Most of the settings are established one at a time with no beam. Two of them, the magnetic field and the source (the filament and gas that make the ions), are explored together. The field sets how fast the ions circle; the RF voltage swings at a fixed rate. Only when the two rates match does an ion reach the gap at the same point of every RF swing, in step with it. Then it gets a push at every crossing. That match is the resonance, and it falls at one field value. So the search is a sweep. With the RF fixed, run the magnet current slowly through the expected field and watch the collector. A peak that comes back at the same field on every pass is the signal. The record of first beams is a record of machines that were far from perfect. The Mullins machine found its first resonances in 2018 with a dee of uncalibrated voltage, an unshielded plate and a filament resting loose on the chamber floor (hosted instrumentation note, §7; the build log's \"It works!\" page). If no peak appears, reconsider the other conditions, the source above all, and repeat.",
   "sources": [
    "dg-089",
    "dg-090",
    "dg-135",
    "dg-1576",
    "dg-1582",
    "dg-1737",
    "dg-302",
    "dg-308",
    "dg-506",
    "dg-514"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/exercises/find-the-resonance/",
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#where-to-ask",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-is-it-working",
    "beam-which-ion",
    "magnet-what-to-measure",
    "ops-the-knobs",
    "ops-what-can-i-test",
    "ops-worked-yesterday",
    "ops-would-my-detector-see-it",
    "rf-which-resonance",
    "source-is-it-making-ions",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/finding-and-keeping-a-beam/#ops-finding-first-beam"
  },
  {
   "id": "ops-worked-yesterday",
   "question": "It worked yesterday. Why doesn't putting every knob back where it was bring the beam back?",
   "group": "G",
   "group_title": "Finding and Keeping a Beam",
   "part": 3,
   "order": 5,
   "tags": [
    "run"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Because the knobs set inputs, and the machine answers with quantities nobody set: the magnetic field rather than the magnet current, the dee voltage rather than the forward power, the pressure at the source rather than the valve turns. Forward power is what the amplifier sends toward the dee. Each drifts for a known reason. The magnet’s iron gives a different field for the same current depending on whether the current rose or fell to it. Warm coils draw less current from a voltage-holding supply. Heat moves the RF resonance as the coil and dee expand. A filament thins as it runs, and surfaces give off gas for hours after the chamber has been open to air. A cyclotron resonance is narrow enough to notice all of them. Orbiting beam exists only where the magnetic field matches the RF frequency. The beam-measurement page’s test for a real peak is that it dies on a small detune, a small move of either off the match. Repeatability comes from recording the answering quantities and returning to them the same way. Cycle the magnet the same way each run and sweep the field slowly enough that a narrow peak cannot slip between readings.",
   "short_answer_markdown": "Because the knobs set inputs, and the machine answers with quantities nobody set: the magnetic field rather than the magnet current, the dee voltage rather than the forward power, the pressure at the source rather than the valve turns. Forward power is what the amplifier sends toward the dee. Each drifts for a known reason. The magnet's iron gives a different field for the same current depending on whether the current rose or fell to it. Warm coils draw less current from a voltage-holding supply. Heat moves the RF resonance as the coil and dee expand. A filament thins as it runs, and surfaces give off gas for hours after the chamber has been open to air. A cyclotron resonance is narrow enough to notice all of them. Orbiting beam exists only where the magnetic field matches the RF frequency. The beam-measurement page's test for a real peak is that it dies on a small detune, a small move of either off the match. Repeatability comes from recording the answering quantities and returning to them the same way. Cycle the magnet the same way each run and sweep the field slowly enough that a narrow peak cannot slip between readings.",
   "sources": [
    "dg-009",
    "dg-090",
    "dg-1428",
    "dg-1737",
    "dg-1847",
    "dg-1858",
    "dg-2064",
    "dg-2075",
    "dg-220",
    "dg-234",
    "dg-241",
    "dg-293",
    "dg-408",
    "dg-456"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/magnet-design/#measurement",
    "https://cyclotroninfo.com/learn/dee-coupling/#measuring",
    "https://cyclotroninfo.com/learn/ion-sources/#arc-behavior",
    "https://cyclotroninfo.com/learn/vacuum/#voltage-holding"
   ],
   "refs": [
    "beam-documenting-the-claim",
    "ops-finding-first-beam",
    "ops-the-knobs"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/finding-and-keeping-a-beam/#ops-worked-yesterday"
  },
  {
   "id": "ops-beam-near-center-not-farther-out",
   "question": "I can detect acceleration near the centre, but not farther out. Is that a partly working cyclotron? How do I find where the particles are being lost?",
   "group": "G",
   "group_title": "Finding and Keeping a Beam",
   "part": 3,
   "order": 6,
   "tags": [
    "run"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Yes, if you have convincingly established that particles are being accelerated at a small radius, with the checks of the previous entries behind the peak: a partly working cyclotron, and a useful one. You have shown that at least some particles can be produced, captured and accelerated far enough to reach that point. You have not shown that they keep surviving and accelerating to larger radii, and the record says that is the ordinary state of a new machine rather than a failure of it. The 63-inch’s first-month currents on a probe, a collector on a movable arm, fell by a factor of two thousand between five inches and its outer radius ( Howard (ed.), p. 10 ). The question is where the ions go. Split the trip into four stages, production, capture, survival and collection, and ask which one the observation belongs to. A verified peak at small radius has shown the first two, for the ions that made it; survival over the many turns outward is the long part. Collection, whether the plate at the larger radius is actually reading what arrives, is checked first: it is the cheapest, and a plate of finite size can miss ions that reach its radius.",
   "short_answer_markdown": "Yes, if you have convincingly established that particles are being accelerated at a small radius, with the checks of the previous entries behind the peak: a partly working cyclotron, and a useful one. You have shown that at least some particles can be produced, captured and accelerated far enough to reach that point. You have not shown that they keep surviving and accelerating to larger radii, and the record says that is the ordinary state of a new machine rather than a failure of it. The 63-inch's first-month currents on a probe, a collector on a movable arm, fell by a factor of two thousand between five inches and its outer radius (dg-802). The question is where the ions go. Split the trip into four stages, production, capture, survival and collection, and ask which one the observation belongs to. A verified peak at small radius has shown the first two, for the ions that made it; survival over the many turns outward is the long part. Collection, whether the plate at the larger radius is actually reading what arrives, is checked first: it is the cheapest, and a plate of finite size can miss ions that reach its radius.",
   "sources": [
    "dg-143",
    "dg-1736",
    "dg-367",
    "dg-494",
    "dg-505",
    "dg-695",
    "dg-790",
    "dg-802",
    "dg-905",
    "dg-909",
    "dg-927"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#probes-shadows",
    "https://cyclotroninfo.com/learn/beam-measurement/#faraday-cup",
    "https://cyclotroninfo.com/learn/ion-sources/#the-job",
    "https://cyclotroninfo.com/learn/beam-dynamics/#healthy-machine",
    "https://cyclotroninfo.com/calculators/path-length/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-is-it-working",
    "beam-what-stops-them-hitting-the-dee",
    "magnet-dc-harmonics",
    "ops-finding-first-beam",
    "ops-would-my-detector-see-it",
    "rf-sub-kilovolt-dees",
    "source-how-many-turns",
    "source-timing-with-rf",
    "source-where-does-it-go",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/finding-and-keeping-a-beam/#ops-beam-near-center-not-farther-out"
  },
  {
   "id": "beam-is-it-working",
   "question": "How do I know my cyclotron is actually working? If I measure a current, how do I know it's accelerated particles rather than something else in the chamber?",
   "group": "H",
   "group_title": "Proving the Beam",
   "part": 3,
   "order": 1,
   "tags": [
    "prove"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "No single reading tells you. The evidence is a collector current that behaves as an accelerated beam must under changes you make on purpose. In resonance the ion’s time round one turn matches one RF cycle, so it is pushed at every gap crossing. For one kind of ion that ties each RF frequency to one magnet field. So hold the frequency, sweep the magnet current and look for a peak at the predicted field. The peak is narrow and goes away when field or frequency is moved a little off it, detuned ( Wouters (1949), p. 11 ). On the Mullins machine at 9.033 MHz the field predicted for a proton is 593 mT (a proton resonates at 15.2 MHz per tesla). The logged peak sat at 596 mT and the sweep went flat with the RF detuned, “consistent with acceleration” (hosted performance note §5, §7, §8). Then run the controls: the same sweep with gas off, with RF off, and with the field detuned well away. Each removes the beam and something else, so each rules out a different explanation, and together they say what the peak must stand above ( How do I know my beam-current detector could detect the beam I'm looking for, before I use it to troubleshoot the cyclotron? ). No single test proves that every other explanation has been excluded.",
   "short_answer_markdown": "No single reading tells you. The evidence is a collector current that behaves as an accelerated beam must under changes you make on purpose. In resonance the ion's time round one turn matches one RF cycle, so it is pushed at every gap crossing. For one kind of ion that ties each RF frequency to one magnet field. So hold the frequency, sweep the magnet current and look for a peak at the predicted field. The peak is narrow and goes away when field or frequency is moved a little off it, detuned (dg-514). On the Mullins machine at 9.033 MHz the field predicted for a proton is 593 mT (a proton resonates at 15.2 MHz per tesla). The logged peak sat at 596 mT and the sweep went flat with the RF detuned, \"consistent with acceleration\" (hosted performance note §5, §7, §8). Then run the controls: the same sweep with gas off, with RF off, and with the field detuned well away. Each removes the beam and something else, so each rules out a different explanation, and together they say what the peak must stand above (ops-would-my-detector-see-it). No single test proves that every other explanation has been excluded.",
   "sources": [
    "dg-1424",
    "dg-1425",
    "dg-1466",
    "dg-1573",
    "dg-1766",
    "dg-1767",
    "dg-1781",
    "dg-1873",
    "dg-506",
    "dg-514",
    "dg-517",
    "dg-801",
    "dg-881",
    "dg-885",
    "dg-927"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/learn/beam-measurement/#faraday-cup",
    "https://cyclotroninfo.com/learn/exercises/find-the-resonance/",
    "https://cyclotroninfo.com/learn/experiments-by-energy/",
    "https://cyclotroninfo.com/learn/safety/",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-documenting-the-claim",
    "beam-which-ion",
    "goal-first-success",
    "ground-loops",
    "ops-finding-first-beam",
    "ops-would-my-detector-see-it",
    "rf-which-resonance",
    "safety-before-beam"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/proving-the-beam/#beam-is-it-working"
  },
  {
   "id": "beam-which-ion",
   "question": "I'm feeding hydrogen into the chamber. Does that mean the accelerated particles must be protons? Why might I see more than one resonance peak?",
   "group": "H",
   "group_title": "Proving the Beam",
   "part": 3,
   "order": 2,
   "tags": [
    "prove"
   ],
   "version": 2,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "No. Hydrogen gas in the source makes three positive ions, H⁺, H₂⁺ and H₃⁺. H⁺ is a bare proton; H₂⁺ and H₃⁺ are two and three hydrogen nuclei bound with one net positive charge. The species has to be inferred from evidence. More than one peak is normal. A heavier ion on the same charge circles more slowly, so staying in step with the RF takes more field: twice the proton’s for H₂⁺, three times for H₃⁺. An ion taking three or five RF cycles per turn is still pushed at each gap crossing, so each species repeats at a third and a fifth of its own field. RF cycles per turn is the harmonic number k; k = 1 is the fundamental. Three times one third is one, so H₃⁺ on k = 3 lands on H⁺ at k = 1 at every frequency, in mass-number arithmetic; the true fields differ by a few hundredths of a percent, less than a peak’s width. A one-frequency sweep reads “hydrogen ions” until other evidence separates them. If the RF allows, lowering the frequency brings the fundamental resonances of several hydrogen ions within your magnet’s range; the scan informs the identification but cannot separate the H⁺ k=1 / H₃⁺ k=3 coincidence. Compute every candidate’s field before naming a peak.",
   "short_answer_markdown": "No. Hydrogen gas in the source makes three positive ions, H⁺, H₂⁺ and H₃⁺. H⁺ is a bare proton; H₂⁺ and H₃⁺ are two and three hydrogen nuclei bound with one net positive charge. The species has to be inferred from evidence. More than one peak is normal. A heavier ion on the same charge circles more slowly, so staying in step with the RF takes more field: twice the proton's for H₂⁺, three times for H₃⁺. An ion taking three or five RF cycles per turn is still pushed at each gap crossing, so each species repeats at a third and a fifth of its own field. RF cycles per turn is the harmonic number k; k = 1 is the fundamental. Three times one third is one, so H₃⁺ on k = 3 lands on H⁺ at k = 1 at every frequency, in mass-number arithmetic; the true fields differ by a few hundredths of a percent, less than a peak's width. A one-frequency sweep reads \"hydrogen ions\" until other evidence separates them. If the RF allows, lowering the frequency brings the fundamental resonances of several hydrogen ions within your magnet's range; the scan informs the identification but cannot separate the H⁺ k=1 / H₃⁺ k=3 coincidence. Compute every candidate's field before naming a peak.",
   "sources": [
    "dg-019",
    "dg-1383",
    "dg-1422",
    "dg-1423",
    "dg-1424",
    "dg-1425",
    "dg-1466",
    "dg-1511",
    "dg-1512",
    "dg-1530",
    "dg-1659",
    "dg-1673",
    "dg-1811",
    "dg-1817",
    "dg-1873",
    "dg-1908",
    "dg-1912",
    "dg-2084",
    "dg-2092",
    "dg-365",
    "dg-502",
    "dg-506",
    "dg-622",
    "dg-801",
    "dg-803",
    "dg-818"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/exercises/weigh-the-ion/",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/learn/ion-sources/",
    "https://cyclotroninfo.com/learn/glossary/#harmonic-number",
    "https://cyclotroninfo.com/learn/glossary/#resonance-condition",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-documenting-the-claim",
    "beam-is-it-working",
    "goal-electrons-or-other-ions",
    "magnet-what-to-measure",
    "rf-which-resonance",
    "source-is-it-making-ions",
    "vac-what-pressure"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/proving-the-beam/#beam-which-ion"
  },
  {
   "id": "beam-which-terms-can-wait",
   "question": "What are all these beam terms, fluence, emittance, dispersion, chromaticity, energy spread and so on? Which do I need to understand now, and which can wait?",
   "group": "H",
   "group_title": "Proving the Beam",
   "part": 3,
   "order": 3,
   "tags": [
    "prove"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "For a first beam you need to understand four things, each something your own instruments show. They are the resonance condition, the current you measure, the background it must stand above, and why one peak may be any of several ion species or harmonics. On a harmonic the ion takes three or five RF cycles per turn, still pushed at each crossing, and peaks at a third or a fifth of the field. The rest of your list describes how a beam is spread out or delivered to a target; it becomes useful when you interpret losses, characterise performance or do experiments, and can wait until then. At amateur energies “good beam” means it survives to full radius, fits inside the dee aperture (the vertical opening inside the dee the beam travels through), and is identifiable from a sharp resonance peak. Emittance, how spread out the particles are in position and direction, “is rarely worth measuring” because the machine is its own transport line: the beam is never taken out into a beamline it would have to fit (beam-quality chapter, #good-beam). The tiers below are a reading order: learn the first now, the others when a problem needs them.",
   "short_answer_markdown": "For a first beam you need to understand four things, each something your own instruments show. They are the resonance condition, the current you measure, the background it must stand above, and why one peak may be any of several ion species or harmonics. On a harmonic the ion takes three or five RF cycles per turn, still pushed at each crossing, and peaks at a third or a fifth of the field. The rest of your list describes how a beam is spread out or delivered to a target; it becomes useful when you interpret losses, characterise performance or do experiments, and can wait until then. At amateur energies \"good beam\" means it survives to full radius, fits inside the dee aperture (the vertical opening inside the dee the beam travels through), and is identifiable from a sharp resonance peak. Emittance, how spread out the particles are in position and direction, \"is rarely worth measuring\" because the machine is its own transport line: the beam is never taken out into a beamline it would have to fit (beam-quality chapter, #good-beam). The tiers below are a reading order: learn the first now, the others when a problem needs them.",
   "sources": [
    "dg-1766",
    "dg-507",
    "dg-521",
    "dg-596"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/glossary/",
    "https://cyclotroninfo.com/learn/beam-quality/#good-beam",
    "https://cyclotroninfo.com/learn/beam-quality/#energy-spread",
    "https://cyclotroninfo.com/learn/beam-quality/#phase-space",
    "https://cyclotroninfo.com/learn/beam-quality/#extent",
    "https://cyclotroninfo.com/learn/beam-measurement/#probes-shadows",
    "https://cyclotroninfo.com/learn/beam-dynamics/",
    "https://cyclotroninfo.com/learn/beam-extraction/",
    "https://cyclotroninfo.com/calculators/cyclotron-energy/",
    "https://cyclotroninfo.com/calculators/path-length/"
   ],
   "refs": [
    "beam-documenting-the-claim",
    "beam-what-stops-them-hitting-the-dee",
    "beam-which-ion",
    "ops-beam-near-center-not-farther-out",
    "ops-would-my-detector-see-it",
    "rf-which-resonance",
    "safety-before-beam",
    "source-timing-with-rf"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/proving-the-beam/#beam-which-terms-can-wait"
  },
  {
   "id": "beam-documenting-the-claim",
   "question": "How should I document that my cyclotron is accelerating protons? What would someone else need to evaluate the claim?",
   "group": "H",
   "group_title": "Proving the Beam",
   "part": 3,
   "order": 4,
   "tags": [
    "prove"
   ],
   "version": 1,
   "revised": "2026-11-01",
   "safety_entry": false,
   "short_answer": "Write down what a stranger would need to repeat your sweep and to doubt it. The record has two levels, and the level sets how strong and specific a claim it supports. It does not divide records into evidence and non-evidence. A useful first-beam record holds the configuration, the RF frequency, the approximate field and how it was read, the collector-current behaviour (trace if you have it), whatever background or detuning checks you could do, and how reproducible it was. It supports “a resonance on this machine consistent with acceleration”; it does not yet support a species or an energy. A stronger, independently reviewable record adds the dated machine description, the operating point, the sweep saved as data, the backgrounds, the checks, the uncertainties and the repeatability, each described further down. A number’s weight comes from how it was obtained, so sort every number into three columns and never let them read alike. Measured: field, frequency, the collector current as the instrument indicated it, pressure, radius. Inferred: the species as candidates, the energy computed from field and radius with its convention, the dee voltage where a method gives one. Designed or expected: the predicted proton field, the design energy and dee voltage. “Accelerating protons” needs the species step, evidence that separates the candidates; until then the phrase is “hydrogen ions”.",
   "short_answer_markdown": "Write down what a stranger would need to repeat your sweep and to doubt it. The record has two levels, and the level sets how strong and specific a claim it supports. It does not divide records into evidence and non-evidence. A useful first-beam record holds the configuration, the RF frequency, the approximate field and how it was read, the collector-current behaviour (trace if you have it), whatever background or detuning checks you could do, and how reproducible it was. It supports \"a resonance on this machine consistent with acceleration\"; it does not yet support a species or an energy. A stronger, independently reviewable record adds the dated machine description, the operating point, the sweep saved as data, the backgrounds, the checks, the uncertainties and the repeatability, each described further down. A number's weight comes from how it was obtained, so sort every number into three columns and never let them read alike. *Measured:* field, frequency, the collector current as the instrument indicated it, pressure, radius. *Inferred:* the species as candidates, the energy computed from field and radius with its convention, the dee voltage where a method gives one. *Designed or expected:* the predicted proton field, the design energy and dee voltage. \"Accelerating protons\" needs the species step, evidence that separates the candidates; until then the phrase is \"hydrogen ions\".",
   "sources": [
    "dg-1517",
    "dg-1581",
    "dg-1652",
    "dg-1766",
    "dg-1811",
    "dg-2052",
    "dg-2057",
    "dg-294",
    "dg-408",
    "dg-883",
    "dg-885"
   ],
   "links": [
    "https://cyclotroninfo.com/learn/beam-measurement/",
    "https://cyclotroninfo.com/learn/beam-measurement/#electrometer",
    "https://cyclotroninfo.com/learn/beam-measurement/#real-beam",
    "https://cyclotroninfo.com/builds/",
    "https://cyclotroninfo.com/learn/so-you-want-to-build-one/#where-to-ask",
    "https://cyclotroninfo.com/library/amateur/#mullins"
   ],
   "refs": [
    "beam-is-it-working",
    "beam-which-ion",
    "goal-first-success",
    "magnet-what-to-measure",
    "ops-the-knobs",
    "ops-worked-yesterday",
    "ops-would-my-detector-see-it",
    "rf-watts-to-volts"
   ],
   "url": "https://cyclotroninfo.com/learn/faq/proving-the-beam/#beam-documenting-the-claim"
  }
 ]
}