Cyclotron controls & instrumentation design rules

gibt es auch bei der Druckmessung kein Messgerät, das über den gesamten Druckbereich von neun Zehnerpotenzen messen kann [tr.: no single gauge covers the nine-decade range]

Prechtl & Wolf, Das Lehr-Zyklotron COLUMBUS — Mit einem Teilchenbeschleuniger Physik und Technik erleben, Springer (2020) — p. 56-57

Editorial note, tabletop extrapolation: Both gauge types are gas-species dependent: a Pirani reads hydrogen differently from nitrogen and a cold cathode needs a hydrogen correction factor, so the pressure that enters the mean-free-path budget should always carry its gas correction; and keep magnetically sensitive heads out of the stray field or shield them per their spec.

Dieser Druck wird durch einen Druckminderer auf pH2 = 300 mbar reduziert [tr.: the pressure is reduced by a regulator to 300 mbar]

Prechtl & Wolf, Das Lehr-Zyklotron COLUMBUS — Mit einem Teilchenbeschleuniger Physik und Technik erleben, Springer (2020) — p. 61

Editorial note, tabletop extrapolation: Fix the reducer pressure and log it - it is a term in the balance. Whatever meters the flow, state its reference conditions and take accuracy and repeatability from its specification rather than assuming a resolution.

Wir legen uns also mit einem geeigneten Detektor auf die Lauer und verändern das Magnetfeld solange, bis wir ein Signal erhalten [tr.: lie in wait with a detector and vary the field until a signal appears]

Prechtl & Wolf, Das Lehr-Zyklotron COLUMBUS — Mit einem Teilchenbeschleuniger Physik und Technik erleben, Springer (2020) — p. 69-71

Editorial note, tabletop extrapolation: Slow ramps help but don't grant immunity: characterize the electrometer/amplifier settling time and pick a sweep rate that resolves the narrowest expected peak - then confirm by comparing up- and down-sweeps (hysteresis and lag shift peaks in opposite directions). Correct the Hall reading to the median plane: a probe in the lid recess reads a different field than the orbit (the 1-7.5 percent class errors below), which moves every q/m assignment.

Ein Steuergerät liefert eine zur Flussdichte proportionale Spannung, die für die weitere Auswertung verwendet wird [tr.: a controller supplies a voltage proportional to B used for evaluation]

Prechtl & Wolf, Das Lehr-Zyklotron COLUMBUS — Mit einem Teilchenbeschleuniger Physik und Technik erleben, Springer (2020) — p. 36, 70

Editorial note, tabletop extrapolation: A fixed probe reads ITS OWN location's field, not the median plane's: map the probe output against a median-plane measurement across the full operating range and both ramp directions (saturation and hysteresis bend the relation), fit the transfer curve, and carry its uncertainty into every specific-charge assignment - a one-point offset calibration is the minimum, not the goal.

ein Wert der in guter Übereinstimmung mit dem gemessenen Druck steht [tr.: a value in good agreement with the measured pressure]

Prechtl & Wolf, Das Lehr-Zyklotron COLUMBUS — Mit einem Teilchenbeschleuniger Physik und Technik erleben, Springer (2020) — p. 53

Editorial note, tabletop extrapolation: One-point agreement checks consistency; to actually MEASURE the pump stand's hydrogen speed at the chamber, do the differential version (baseline, step the flow, gauge-corrected delta-p) and repeat after every plumbing change or pump swap, logging the result.

The oscillator used an HP8165 digital programmable RF signal source, which offered an easy method of tuning the frequency. ... [Later the] signal was produced using a HP8656B signal source, which had greater frequency resolution than the HP8165.

Cressman, The Design and Construction of a Small Cyclotron — Houghton College thesis (2006) — p. 17

Editorial note, tabletop extrapolation: Resonance hunting rewards fine, repeatable frequency steps - the documented upgrade was FOR resolution, so check any candidate source's step size against the measured resonance width (Q of the loaded resonator) before buying; whether the first unit's resolution actually limited tuning is our inference from the upgrade, not the thesis's statement.

an interlock which shuts down the magnetic if less than 2.5 liters per minute of chilled water are supplied

Cressman, The Design and Construction of a Small Cyclotron — Houghton College thesis (2006) — p. 36

Editorial note, tabletop extrapolation: The transferable pattern is the method, not the numbers: independent low-flow and over-temperature interlocks wired to POWER DOWN the load, with trip points derived from the coil's insulation limits or measured thermal performance (including sensor lag) - Houghton's 2.5 L/min floor and their coil ceiling are that machine's settings, not defaults.

in a concrete brick room with an interlock control system to prevent the accelerator from being turned on when a person is in the room

Cressman, The Design and Construction of a Small Cyclotron — Houghton College thesis (2006) — p. 29

Editorial note, tabletop extrapolation: The documented access-control arrangement of the source machine - occupancy interlock plus (per its electronics chapter) remote operation - is a COMPONENT of protection, not a certified minimum: shielding calculations, surveys, monitors, fail-safe interlock design and applicable regulatory requirements decide sufficiency for any neutron-capable machine, and the thesis presents no dose analysis.

All of the electronics, with the exception of the floating filament power supply, are monitored and controlled remotely through the general purpose interface bus. The National Instruments GPIB-enet allows these instruments to be controlled through an Ethernet network. ... The 1-100-K Ion gauge and CVT-272-101 Convectron gauge are connected to an SRS FGC 100 Ion Gauge Controller ... The SRS RGA 100 connection is RS232, and so it needs the National Instruments RS232-GPIB Converter ... The Powerten R62B-4050 magnet power supply supports a GPIB connection ... The voltage on the filament floats on the voltage provided by 0-100V power supply

Cressman, The Design and Construction of a Small Cyclotron — Houghton College thesis (2006) — p. 47

Editorial note, tabletop extrapolation: Full remote operation is what makes an occupancy interlock workable. The floating-filament exception carries the real lesson: a floated circuit must not connect directly to ground-referenced instrumentation - it needs an isolated interface (or manual presetting outside the run), which is an implementation choice, not an impossibility.

The power in the primary coil of the transmatch is monitored by a Bird 43A RF power meter, and supplied by the ENI 155LCRH RF power amplifier. The RF signal is provided by the HP 33120A function [generator]

Cressman, The Design and Construction of a Small Cyclotron — Houghton College thesis (2006) — p. 47

Editorial note, tabletop extrapolation: A bench function generator + lab RF amplifier + ham-style through-line wattmeter is a complete drive-and-monitor chain from commodity gear. Meter honestly: a directional wattmeter at the transmatch primary reads forward power at that point - net delivered power is forward minus reflected, and network losses sit downstream of the meter, so pair it with the dee-voltage pickup (dg-1392's two-readout rule).

the detector is set to a specific position and the magnetic field is continuously increased. With very specific magnetic fields, there are peaks in the beam current

Wolf & Prechtl, COLUMBUS — A Small Cyclotron for School and Teaching Purposes — THPO001, Proceedings of Cyclotrons2022 (2022) — p. 2

Editorial note, tabletop extrapolation: A B-sweep at fixed frequency is a q/m RESONANCE SURVEY - the cheapest species diagnostic a small machine has, not a full mass spectrometer: state the harmonic number, calibrate the field reading (dg-1428), and resolve the q/m degeneracies and harmonic ambiguities by field-ratio checks or frequency scaling (dg-1422/dg-1423) before naming species.

A directional-coupler in the input-circuit of the matchbox makes it possible to control and minimize the reflections back into the RF-source and a RF pick-up, i.e. Fig. 6, in the output-circuit allows to check whether the cyclotron is tuned to the cyclotron-frequency of 5.63 MHz

Wolf, Frank & Held, COLUMBUS — A Small Cyclotron for School and Teaching Purposes — WE1PB03, Proceedings of Cyclotrons2013 (2013) — p. 2

Editorial note, tabletop extrapolation: Two independent indications, reflected power at the input and detected RF at the dee side, help separate matching problems from resonance problems during tune-up - though both respond to coupling and resonance, so neither is unambiguous alone, and the pick-up reads amplitude: the drive frequency itself should be known independently (a counter is cheap) and compared against qB/2πm.

Thermal cut-off switches (fully insulated in a screw on housing type), set to open an electrical circuit at 40°±5°C shall be fitted on the external lead (return lead of water circuit) of each pancake ... Suitable thermal switches will be placed on outer terminals of the coils to prevent over-heating of the coils (cut-off value: > 40 oC) by shutting off the power supply ... [spec table:] Thermal sensors (cut-off value) — > 400 C (8 nos. of sensors to be mounted on each coil); Interlocks — overload, high temperature cut-off

IUAC, e-Tender 09/GOR/2024–25 — H-Dipole Water-Cooled DC Electromagnet for the Table-Top Cyclotron: Engineering Specification and Acceptance Tests (2024) — p. 19, 24, 25

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: hardware thermal switches on every pancake's return water lead, dropping the supply through an interlock, is a simple, software-free protection pattern for a water-cooled coil stack — with two cautions. An outlet-mounted switch lags stagnant-water and winding hot spots when flow is lost, so pair it with flow or pressure detection; and the interlock must command the supply's controlled shutdown or energy-dump path, never break magnet current mechanically — an inductive circuit interrupted dry arcs.

Stability of power supply, at least 100 ppm ... [compliance sheet:] Equipment required for field mapping: a) 3D magnet field mapping system with Hall probe, associated control software for the field mapping ... DC Power supply rating: Voltage: 20V, Current: 200 Amps ... CMM and allied measuring instruments ... In-house electrical testing facilities: Insulation Resistance, Hipot Test, Inductance ... Inhouse Hydraulic Testing Facility for Coils ... Note: Final Testing of assembled magnet will be performed at supplier premises, hence supplier is required to provide a list of testing facility available in-house.

IUAC, e-Tender 09/GOR/2024–25 — H-Dipole Water-Cooled DC Electromagnet for the Table-Top Cyclotron: Engineering Specification and Acceptance Tests (2024) — p. 40, 41

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: the equipment list doubles as a checklist of what a serious small-magnet test stand contains, and 100 ppm shows what a professional team asks of a mapping/energizing supply. It is one machine's specification, not a universal requirement: derive the allowable current stability from the machine's own B-I slope, field tolerance and RF phase-slip budget — the answer is usually far tighter than an unregulated bench supply but need not be 100 ppm.

IUAC, New Delhi, India, operates accelerators with RF structures in the range of 12.125-97 MHz, in both normal and superconducting modes ... this controller has been tested as a Sawtooth Waveform Generator for the Multi-Harmonic Buncher (MHB), as a generator-driven (GDR), and as a self-excited loop (SEL)-based LLRF for various RF cavities at IUAC ... It is a compact, frequency-reconfigurable, standalone device controlled by an EPICS IOC ... The main feature of our design is the hardware configuration, which remains the same regardless of the cavity type, without the need for FPGA reprogramming. In addition to the LLRF algorithm, the same FPGA contains logic for the motorized Frequency Tuner Control, considerably lowering the system's cost ... The major blocks of the system as shown in the overall physical block diagram (Fig. 1a) are a wideband Analog Front-End (AFE), microcontroller programmed PLL Multiplier, and a SoC-FPGA-based digital board.

Sharma, Kar, Dabas, Sahu, Singh, Mathur, Satyanarayana & Venkatramanan, Design and Testing of a Universal Embedded Feedback Controller for RF Cavities — MOP6698, Proceedings of IPAC2026 (2026) — p. 1, 2

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: one reconfigurable digital controller replacing a zoo of structure-specific analog LLRF chassis is exactly the maintainability trade a small lab faces. Folding the mechanical tuner drive into the same FPGA as the feedback loop removes a separate controller box — the motor's power stage still exists, but its logic does not need its own electronics.

Long-term RMS stability of ~1% in amplitude and a < ±0.4∘ in phase locks have been obtained ... [Table 1, long-term performance:] MHB-DPLL — phase ± 0.40∘; GDR — ± 0.5%, ± 0.45∘; SEL-AP — ± 1.2%, ± 0.35∘ ... The loop allows phase corrections of up to 35∘ and amplitude corrections of ±3 dB, as verified using an additional phase shifter and attenuator.

Sharma, Kar, Dabas, Sahu, Singh, Mathur, Satyanarayana & Venkatramanan, Design and Testing of a Universal Embedded Feedback Controller for RF Cavities — MOP6698, Proceedings of IPAC2026 (2026) — p. 1, 3

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: sets a benchmark for what percent/sub-degree RF regulation looks like from a compact digital controller — and, more transferably, shows how to verify a loop's correction range by deliberately injecting known phase and amplitude disturbances. Bench figures of this style are the right pre-beam acceptance evidence for any home-built dee drive.

this mode also features a PWM-controlled motorized frequency tuner in the same FPGA ... It compares the phase error (between the FWD signal and the PU signal) with a threshold value and, based on that, it decides the direction of motion of the tuner

Sharma, Kar, Dabas, Sahu, Singh, Mathur, Satyanarayana & Venkatramanan, Design and Testing of a Universal Embedded Feedback Controller for RF Cavities — MOP6698, Proceedings of IPAC2026 (2026) — p. 2

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: the forward-versus-pickup phase comparison is the classic resonance-tracking criterion, stated here in an implementable threshold-and-direction form suitable for a microcontroller and a stepper on a trimmer capacitor. Separating slow mechanical tuning from the fast electronic loop is the standard division of labour worth copying — and the on-resonance phase setpoint must be calibrated for the actual coupling, pickup placement and cable delays, not assumed to be zero.

Several implementations invoke backplane-based methods, such as VME, cPCI, and microTCA. These solutions are often costly, bulky, and difficult to adopt due to a customized design goal. Standalone, non-crate-based systems provide a more versatile, fast, and cost-effective alternative ... The local oscillator signal for the mixer is generated by a Si5356-based I2C-programmable PLL multiplier, synchronized with an external reference signal. Apart from the LO signal, it is used to generate a 125 MHz system clock signal for the FPGA ... This board [13] houses the main signal processing algorithm ... [reference 13:] Red Pitaya, "Red pitaya STEMlab 125-14" ... A lightweight digital PLL (DPLL) ... helps the Si5356-based PLL multiplier mitigate long-term sub-millihertz-level errors which were experimentally observed and caused by accuracy issues with the manual setting of its phase increment word

Sharma, Kar, Dabas, Sahu, Singh, Mathur, Satyanarayana & Venkatramanan, Design and Testing of a Universal Embedded Feedback Controller for RF Cavities — MOP6698, Proceedings of IPAC2026 (2026) — p. 1, 2, 4

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: a national accelerator centre building its cyclotron LLRF around a hobbyist-priced software-defined-radio board shows amateur-accessible hardware can anchor serious RF field control at these frequencies — as one component of a system whose performance also hangs on the analog front end, clock reference, firmware and interlocks. The DPLL fix for the clock chip's long-term drift (sub-millihertz-level, from manual phase-increment setting) is a practical gotcha worth knowing before trusting a cheap synthesizer unsupervised.

Frequency is fine tunes with vacuum variable capacitor. A capacitive pick-up built-in the Cyclotron chamber is used as feedback in order to control and maintain the RF voltage and frequency of the system using a digital LLRF controller and a motorized tuner.

IUAC, Annual Report 2024–25, Chapter 3 — Research Support Facilities (table-top cyclotron RF system) — p. 18

Editorial note, tabletop extrapolation: Editorial note, tabletop extrapolation: building the capacitive pick-up into the chamber from the start — rather than improvising one later — is the design habit to copy: early provision simplifies every scheme that reads the cavity field from a pick-up, including the dee-voltage calibration chain on this machine. Other feedback routes exist (directional-coupler signals, other probe types); a motor-driven vacuum variable capacitor is an amateur-accessible tuner implementation.

The active tuning system must be designed to bring the cavities on the frequency initially to compensate detuning for temperature variations due to RF heating and can be realized by capacitance tuner from radial direction.

Karamyshev, A New Design of PET Cyclotron — IPAC2022, paper THPOMS016, JACoW (2022) — p. 3

Editorial note, tabletop extrapolation: Thermal detuning follows the power actually dissipated in the resonator walls and where it lands, not the drive power alone; it is smaller at the dee voltages tabletop machines run but not zero, and a high-Q resonator still drifts as dee and stem warm. Estimate the wall loss and the resulting frequency shift against the resonator's bandwidth; a trimmer capacitor or a frequency-following drive serves the same function as the paper's radial tuner.

When the current is increased, therefore, the total positive ionization of the beam increases, the field around the beam becomes stronger, and the electrons are brought to a focus in a shorter distance.

Johnson, A Low Voltage Cathode Ray Oscillograph — Bell System Technical Journal 1(2), 142–153 (1922) — p. 7

Editorial note, tabletop extrapolation: In Johnson's ion-focused tube raising the current shortened the focal distance, so intensity and focus may be coupled in similar gas-filled low-energy beamlines: an emission-current change can move the waist. If focus and intensity must be set separately, provide an independent focusing element (a magnetic or electrostatic lens, or another characterised focusing parameter) — an independent current control alone does not decouple them.

The control system has function of resonance frequency tracking of the cyclotron during operation and adjusts simultaneously the operation frequency of the generator.

Smirnov, Ivanov, Krasnov, Polikhov, Rezanov & Sharkov (Siemens Research Center, Moscow), 72 MHz Solid-State Amplifier Power Test — IPAC2014, paper WEPME008, JACoW (2014) — p. 1

Editorial note, tabletop extrapolation: The telemetry transfers to a tabletop drive: drain voltage/current, drive level and forward/reflected power per stage are cheap to measure and are the inputs any protection loop needs. The frequency tracking does not transfer as stated: in a fixed-field cyclotron the RF frequency is set by field and harmonic for beam synchronism, so the dee is held on resonance by a mechanical or electrical tuner while the RF stays at the beam-synchronous frequency — letting the generator chase cavity drift would slip the beam phase.

The signals from the RF pickup loops, including spare signals, are digitized by AD converters and further processed on a FPGA board. Beside the basic feedback functionality also high level functions for pulse processing and machine protection are integrated. The flexibility in re-programming of the FPGA allowed fast optimization of machine protection procedures and implementation of new operational features in order to improve the overall RF system stability.

Röcken et al. (Varian Medical Systems Particle Therapy), Progress at Varian's Superconducting Cyclotrons: A Base for the ProBeam Platform — Cyclotrons 2013, paper MOPPT012, JACoW (2013) — p. 2

Editorial note, tabletop extrapolation: The architecture, not Varian's hardware, is what applies to a tabletop drive: direct digitisation of a dee pickup is within reach of low-cost FPGA or SDR boards (an engineering observation, not the paper's), and putting reflected-power and pickup-collapse trips in the same re-programmable logic as the amplitude loop lets protection be tuned during conditioning — with independent fail-safe interlocks kept in place.

The SSA is still protected by a second unfiltered fast interlock that triggers if the reflected power transient exceeds 150 kW.

Jacob (ESRF), Radio Frequency Solid State Amplifiers — CAS–CERN Accelerator School: Power Converters, CERN-2015-003, pp. 197–216 (2015); arXiv:1607.01570 — p. 18

Editorial note, tabletop extrapolation: The two-tier logic transfers to a tabletop LDMOS dee drive as one mitigation: a slow, averaged reflected-power trip tolerates the fill and switch-off transients of soft-start or pulsed conditioning, while a fast peak trip catches arcs. Thresholds and response times come from worst-case limits for the whole chain — the device's ratings under the specified bias, phase, pulse and thermal conditions, plus combiner, match, load and detector latency — across all mismatch phases, validated at reduced power before staged qualification, with direct arc detection where appropriate.

As such, the system allows for operating frequency selection from 49 to 80 MHz and is compatible with single or double resonator configurations through the use of local oscillator synchronization and high-speed command exchange. An I/Q demodulation/modulation scheme is employed allowing for frequency and amplitude control. […] The FPGA is responsible for the high-speed signal processing such as the amplitude control loop, spark recovery, interlocks and safety. […] The Analog Devices Blackfin microcontroller handles the higher level interface with the motor drivers and Host PC. […] The operational frequency of the boards can be selected for a cyclotron between 49-80MHz by setting a few switches and restarting the LLRF. As this change does not require any hardware or software modifications it greatly increases the utility and modularity of the LLRF controllers.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 1

Editorial note, tabletop extrapolation: The fast/slow split transfers to a tabletop dee drive: spark response and amplitude regulation belong in programmable logic or a dedicated fast loop, while tuner motors and sequencing tolerate a microcontroller or host. Frequency selectability by switches is this 49–80 MHz design's; at a few MHz a direct-sampling digitiser can replace the IF mixing stage — an engineering option the paper does not address.

1. Pulse the RF power at 20% of full power until the cavity readback is above 15% of full power while tuning 2. Transition from pulsed RF to CW and enable the spark recovery and frequency tuning loop 3. Close the amplitude control loop 4. Ramp up the RF drive to full power […] All of the timing and threshold parameters are digitally adjustable which allows for rapid tuning of the procedure to ensure reliable start-up and operation.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 2

Editorial note, tabletop extrapolation: The paper states the aim as minimising the time from RF off to full power. As an engineering reading (not stated in the paper), a pulsed first step limits the average power into a resonator that is still off tune only to the extent its duty factor is below unity — the paper does not give the duty factor — and lets the tuner find resonance before CW is applied; reflected-power protection rests on measured mismatch and the amplifier's ratings, not on the pulsing. The 20 %/15 % numbers are the Best system's settings; for a different resonator they are starting values, which is why the source makes them adjustable.

Sparks are detected by monitoring for a rapid decline in the cavity voltage which happens on the order of 1µs. After detection the recovery procedure pauses the PI amplitude loop, turns off the RF drive for 2µs, applies an open loop drive to regain cavity voltage and then resumes the amplitude control. Properly regulated cavity voltage is restored within approximately 100µs for a typical case. During the cavity voltage recovery, the spark detection is again enabled to protect against the case where an additional spark occurs during the ramp.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 2

Editorial note, tabletop extrapolation: Pausing the integrator during the event is the key transferable step: a PI loop left running sees a collapsed read-back and winds up to full drive, which then lands on the arc. The microsecond timings are the 49–80 MHz Best resonators'; a resonator's own field-amplitude time constant scales as 2Q_L/omega, so a lower-frequency, lower-Q dee sets its own detection and refill times, to be measured on that dee.

The loop is implemented in the FPGA using a hardware description language, and also includes overshoot protection and integral anti-windup. An initial cutoff frequency of 20kHz is selected and can be fine tuned during the particular cyclotron commissioning.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 2

Editorial note, tabletop extrapolation: Anti-windup limits integrator accumulation during saturation (start-up, sparks, tuner moves) and an overshoot limiter bounds commanded-voltage excursions — control hygiene that transfers regardless of frequency, and that reduces the loop's own contribution to stressing a solid-state stage or a marginal dee gap. Independent, fail-safe arc, overvoltage, overcurrent and reflected-power interlocks are still required and validated on the actual RF plant. The 20 kHz figure is an initial setting for Best's resonators; the usable loop bandwidth on another dee is constrained by that resonator's own pole at its half-bandwidth f0/(2Q_L) (general resonator physics, not from the paper) and has to be set on the actual dee.

Automatic frequency control logic is contained in the microcontroller and has a response time of approximately 100Hz. The loop acts to maintain a constant phase offset between the forward drive and cavity voltage. Motor position feedback is given by a linear resistor and optical encoder. Frequency control is available in pulsed and CW mode.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 2

Editorial note, tabletop extrapolation: Forward-versus-pickup phase is the standard tuning error signal and is directly usable on a tabletop dee with a motor-driven trim capacitor, once calibrated for cable delay, coupling and beam loading; keeping the tuner loop much slower than the amplitude loop avoids the two fighting. The source's 'response time of approximately 100Hz' does not define its metric (bandwidth, update rate or something else), so it is not a design specification until the implementation or a measured loop response says otherwise.

This can produce either very short pulses for specific applications, or relatively long pulses with an adjustable duty factor to decrease the overall beam intensity during commissioning.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 2

Editorial note, tabletop extrapolation: Phase-lock is needed where resonators are independently driven and resonant; a shared amplifier feeding separate resonators does not by itself fix the gap-voltage phases (detuning, loading and unequal feed paths move them), while a single-dee machine has no second phase to lock. Programmed relative-phase gating for pulsing is distinct from a static phase offset, which changes acceleration and can raise internal beam loss — if phase is used as an intensity knob during commissioning, measure delivered beam and internal losses.

Fine tuning of the amplitude control loop is continuing to reach a stability target of 5 · 10−4 or better.

Gold & Sabaiduc (Best Cyclotron Systems), Design of a Digital Low-Level RF System for Best Medical Cyclotrons — Cyclotrons 2013, paper TUPPT024, JACoW (2013) — p. 3

Editorial note, tabletop extrapolation: 5 x 10^-4 is a stated goal for Best's resonators, not a demonstrated figure (Fig. 4 axes are unscaled), so it serves only as a reference point for what a commercial system aims at. An hours-long, initially supervised commissioning soak with spark recovery active is the transferable test for a dee drive; unattended operation comes only after independent protective trips (arc, vacuum, reflected power, thermal) and bounded restart logic have been validated.

For a synchronized operation of the modules, a distribution of the reference frequency and a 10 MHz clock signal had to be prepared using existing RF splitters.

Fanselow, Bundesmann, Denker et al. (Helmholtz-Zentrum Berlin), Operational Experience in the Treatment of Ocular Melanomas with a New Digital Low-Level RF Control System — Cyclotrons 2019, paper TUP007, JACoW (2019) — p. 1

Editorial note, tabletop extrapolation: A single shared reference and clock is what lets separate controllers (dee drive, buncher, any second dee) be synchronised; fixed relative phase after start-up additionally needs deterministic phase initialisation or a closed-loop phase measurement. For one dee and one controller it reduces to locking the controller to a stable reference. The HZB RF runs at 10–20 MHz with two 30 kW tube amplifiers (2 kW semiconductor drivers); the reference-distribution scheme is independent of amplifier power.

In order to ensure a fast connection of the analogue, RF, control and interlock signals of the new RF system and a fast change between the old and the new RF control, all signals of both RF systems were connected on patch panels (Fig. 3). Subsequently, all analogue and RF signals were measured at the patch panels under different operating conditions of the cyclotron RF. Amplifiers and attenuators were prepared to adjust the level differences between the measured and the required signal levels of the LLRF. […] In total, only one week was needed for commissioning instead of the planned two weeks. […] Commissioning was faster than planned, with the installed patch panels and previous testing of the motion control modules proving essential.

Fanselow, Bundesmann, Denker et al. (Helmholtz-Zentrum Berlin), Operational Experience in the Treatment of Ocular Melanomas with a New Digital Low-Level RF Control System — Cyclotrons 2019, paper TUP007, JACoW (2019) — p. 4

Editorial note, tabletop extrapolation: The transferable practice is to characterise every pickup and forward/reflected signal level across the operating conditions before connecting a new controller, and to wire them so that controllers can be swapped without re-terminating. At small scale the same discipline applies to moving from a hand-tuned signal generator to a closed-loop controller.

The RF systems were tuned with the old RF control to the standard therapy frequency of 19.3187 MHz to obtain a starting value of the resonance tuning for the LLRF.

Fanselow, Bundesmann, Denker et al. (Helmholtz-Zentrum Berlin), Operational Experience in the Treatment of Ocular Melanomas with a New Digital Low-Level RF Control System — Cyclotrons 2019, paper TUP007, JACoW (2019) — p. 2

Editorial note, tabletop extrapolation: Starting a tuner servo from a recorded, known-resonant tuner position keeps first closed-loop operation near resonance; verify resonance and matching at reduced power while watching reflected power and the protective trips before raising drive. The 115 kV amplitude is HZB's, far above a tabletop dee's few kilovolts — compare by the method, not the numbers.

Already during the adjustment of the accelerator the starting of the RF systems by the automatic resonance search of the LLRF was easier and faster than the manual resonance search of the old system. Thus, the cold start of the RF systems after the maintenance period can be carried out by all operators themselves without any problems and without the help of the RF group.

Fanselow, Bundesmann, Denker et al. (Helmholtz-Zentrum Berlin), Operational Experience in the Treatment of Ocular Melanomas with a New Digital Low-Level RF Control System — Cyclotrons 2019, paper TUP007, JACoW (2019) — p. 3

Editorial note, tabletop extrapolation: Automating the find-resonance step (a low-power sweep or tuner search, then lock) is an optional feature worth building into a small machine's first controller, since its resonance drifts with temperature between sessions; its sweep limits, reflected-power protection during the search and lock acquisition have to be validated.

Both systems now run more stable, with phase stability improving from 0.1° to 0.02° and amplitude stability from 0.5% to 0.7%.

Fanselow, Bundesmann, Denker et al. (Helmholtz-Zentrum Berlin), Operational Experience in the Treatment of Ocular Melanomas with a New Digital Low-Level RF Control System — Cyclotrons 2019, paper TUP007, JACoW (2019) — p. 3

Editorial note, tabletop extrapolation: These are HZB's numbers for its resonators and its definition of stability (not stated in the paper), useful as a reference for what a digital LLRF achieves at 10-20 MHz, not as a requirement for a different machine. The amplitude pair is reported as printed; whether it is a misprint or a change of measurement basis cannot be resolved from the paper.

the problem was identified as a suboptimal tuning of the RF power amplifiers output stage, resulting in the loop instability. To solve the problem, the integral coefficient of the phase modulator’s PID controller was first lowered. This allowed the power amplifier to be optimised at rated power. Thereafter the integral coefficient was returned to its nominal value and normal operation continued.

Fanselow, Bundesmann, Denker et al. (Helmholtz-Zentrum Berlin), Operational Experience in the Treatment of Ocular Melanomas with a New Digital Low-Level RF Control System — Cyclotrons 2019, paper TUP007, JACoW (2019) — p. 3

Editorial note, tabletop extrapolation: The lesson plausibly transfers across frequency and power: an LLRF loop's stability depends on the complex gain and phase response of the power stage it drives, so a re-tuned or mistuned amplifier output network can destabilise a loop that was stable before. When a loop goes unstable after a frequency change, check the amplifier's transfer response, loading and output-network tuning before re-tuning the controller.

Control of the arc-type ion source used in the calutron is normally obtained by use of a "filament-arc regulator" designed to permit independent adjustment of arc current and arc voltage. The filament is operated emission-limited with the limiting emission and arc current being manually adjustable by variations of the filament current. The arc voltage is also manually adjustable, but changes in arc current that would normally accompany arc voltage changes are prevented by a filament-arc regulator which automatically introduces a compensating variation in the filament current.

Yonts, Filament Life in Low Pressure Arc Type Ion Source — ORNL-1758, Oak Ridge National Laboratory (1954) — p. 12

Editorial note, tabletop extrapolation: ORNL-1525 (dg-2079) places 1950s cyclotron filament sources in this same emission-limited control family (arc current set by filament temperature). Yonts's iron-separation runs with the regulator bypassed (dg-2071) are his test of its cost: a regulator that answers every disturbance by changing filament temperature can drive the filament hotter than it needs to be.

Filament life was almost doubled (to 25-30 hours for 170-mil filaments) by this mode of operation

Yonts, Filament Life in Low Pressure Arc Type Ion Source — ORNL-1758, Oak Ridge National Laboratory (1954) — p. 15

Editorial note, tabletop extrapolation: In Yonts's iron-calutron conditions, bypassing the regulator as part of this operating change (manual control at essentially constant filament current, a series ballast, arc voltage set by pressure) nearly doubled life. For a filament source on a feedback loop, check whether the loop is moving filament temperature to chase arc or beam current; fixed filament current — not fixed temperature, since resistance and heat loss change in service — may lengthen life at the cost of output steadiness, which is a test to run, not a guaranteed trade. The doubling is a calutron result on one charge material.

These decay times are inversely proportional to the average filament erosion rates during the selected period of operation

Yonts, Filament Life in Low Pressure Arc Type Ion Source — ORNL-1758, Oak Ridge National Laboratory (1954) — p. 29

Editorial note, tabletop extrapolation: A logged filament current (or resistance, from current and voltage) is an inexpensive wear gauge for a characterized filament run from a stable, specified heater supply under otherwise constant conditions: with the heater at fixed voltage the current falls as the wire thins. A trend indicates wear; converting its slope to an erosion rate needs calibration (for instance against weighed filaments) and exclusion of thermal and electrical causes such as temperature, resistivity, supply regulation and contact resistance. The calutron figures (4.8 A to 3.0 A transformer-primary current) are specific to that supply.

The normal operation of many arc-type ion sources (in particular, various high-current calutron and cyclotron sources) involves control of the arc by changing filament temperature. This type of control is generally used because it permits independent adjustment of arc voltage and current (Fig. 1).

Yonts, Filament Life Extension in High-Intensity Ion Sources — ORNL-1525, Oak Ridge National Laboratory (1953) — p. 5

Editorial note, tabletop extrapolation: The statement places 1950s cyclotron filament sources in the same control family as calutron sources, which motivates checking a cyclotron filament source for an analogous erosion/output trade-off; the calutron's quantitative result need not carry over, since gas species, sheath voltage, cathode material and ion flux change it. It applies to emission-limited filament arcs; a source run without an arc (pure electron impact) has no arc voltage to trade.

of four 150-watt incandescent bulbs in parallel was inserted in series with the arc supply. Arc current was adjusted by changing the supply voltage by means of the induction regulator normally used to change arc voltage. Filament temperature was adjusted by means of the maximum-filament-current control mentioned previously.

Yonts, Filament Life Extension in High-Intensity Ion Sources — ORNL-1525, Oak Ridge National Laboratory (1953) — p. 6

Editorial note, tabletop extrapolation: A series ballast lets the arc find its own voltage at a current set jointly by the supply voltage and the ballast's nonlinear load line, which is one way to try low-voltage (diode-like) operation on a filament arc without a regulated supply. Lamps as ballast have a resistance that rises with current; that behaviour, and the bulb rating needed, depend on the arc current of the source in question.

For example, the control unit 8 may change the direction of the current on the basis of the detection state of ions extracted from the chimney 10.

Miyashita & Matsumura (Sumitomo Heavy Industries), Cyclotron — US Patent 9,041,318 B2 (2015) — p. 12

Editorial note, tabletop extrapolation: The patent tests DC reversal against fixed DC only, and Sumitomo's own 2019 centre-region paper (Kamiguchi et al., dg-2099) reports AC heating adopted on its newer source, also without a comparison. The Handbook of Ion Sources notes that AC heating increases plasma instabilities and the energy spread of the extracted ions (dg-381), a source-specific observation rather than a universal one. A logged reversal schedule lets beam-current changes at each reversal be checked as part of the source record.

Minor adjustments in the magnetic field will be made by using 'field tuners,' movable magnets which can be oriented such that their magnetic field is parallel, antiparallel, or any angle with respect to the existing field.

Li, Cooper, Halbach, Kunkel, Leung et al. (Lawrence Berkeley Laboratory), A Permanent Magnet System for a Cyclotron Used as a Mass Spectrometer — LBL-32037; Cyclotrons 1992, paper XB-02 (1992) — p. 5

Editorial note, tabletop extrapolation: This is the mechanism that answers the fixed-field objection to a permanent-magnet cyclotron: a small, mechanically rotated fraction of the magnet material can trim the field toward resonance at an RF frequency near the nominal design value, provided the required correction lies within the measured trim authority and preserves acceptable field uniformity. The paper is a design (instrument under construction in 1992); the tuning range the rods achieve is not reported here, so the trim authority for a different magnet has to be computed or measured.

Cited in: Cyclotron Magnet Design

The central field can be continuously varied from 0.07 T to 0.3 T by adjusting the location of an iron shunt. […] This shunt has tapered tips which become saturated as the shunt is moved towards the magnet, limiting the amount of flux that can be shunted in proportion to the cross section of the saturated tips. […] The C-shaped shunt yields a much more linear tuning curve and greatly extends the range of travel. This not only relaxes the precision with which the shunt must be moved but also reduces the magnetic force on the shunt

Barlow, Kraus & Meyer (Los Alamos National Laboratory), Variable-Field Permanent-Magnet Dipole — LA-UR-93-3237 (1993) — p. 5

Editorial note, tabletop extrapolation: A movable iron shunt is a second route (besides rotating tuner magnets) to a permanent-magnet field that can be set without coil current. The ~4:1 range is this dipole's; a cyclotron trim range is set by its own resonance-search and drift needs, a shunt for that purpose would be sized for that range, and it must preserve the azimuthal symmetry and radial profile the orbits need. Shunt shaping is a lever for making the setting resolution more uniform across the range.

Cited in: Cyclotron Magnet Design

All linkages in the shunt drive system are designed for minimum backlash, although this is not strictly required since the drive system is always pre loaded against the pull of the shunt.

Barlow, Kraus & Meyer (Los Alamos National Laboratory), Variable-Field Permanent-Magnet Dipole — LA-UR-93-3237 (1993) — p. 5

Editorial note, tabletop extrapolation: A verified unidirectional magnetic load preloaded this particular mechanism. For another movable iron or magnet element, calculate or measure the force or torque over the full travel (it can vanish or reverse), provide power-off holding or a fail-safe where needed (the brake here), and treat limit switches plus step counting as a home datum with incremental tracking that can be lost after slip or power loss — use explicit homing or a true absolute encoder. The same considerations apply to rotating tuner rods, which feel torque from the main field.

Cited in: Cyclotron Magnet Design

In this hybrid magnet, the dotted material is soft iron. The field distribution is controlled by the iron surface, thus it is very insensitive to material properties as long as the permeability of the iron is large compared to 1. […] By rotating the outer ring with the attached CSEM one can clearly change the field strength in the central region.

Halbach (Lawrence Berkeley Laboratory), Use of Permanent Magnets in Accelerator Technology: Present and Future — LBL-23556 (1987) — p. 9

Editorial note, tabletop extrapolation: Two candidate ideas for a permanent-magnet cyclotron: the iron pole face carries the field-shape accuracy, and field strength might be varied by rotating magnet material behind or outside the pole rather than by coils. Halbach states them for a quadrupole; applying them to a cyclotron dipole requires 3-D field analysis and measurement to verify average field, radial profile, symmetry and isochronism, and the insensitivity holds only while the iron is unsaturated.

Cited in: Cyclotron Magnet Design

Given the day to day and seasonal variation of the tunnel, it was impossible to achieve this with compensator strips alone. The 40-inch quads use electric blankets with a feedback loop to maintain the magnets at 32 ± 1 oC.

Brown et al. (Fermi National Accelerator Laboratory), Time Dependence and Temperature Stability of the Permanent Magnets for the Fermilab Antiproton Recycler Ring — FERMILAB-Conf-99/267 (1999) — p. 4

Editorial note, tabletop extrapolation: Heating a PM magnet to a regulated set point (a heater-only loop implies a set point above the warmest ambient, an inference not stated by the source) turns a passive drift into a regulated one; how well it holds depends on the heater, insulation and sensor placement of the actual build and has to be measured on it. The Recycler pre-warmed each ring quadrupole overnight to 32 degC before final adjustment, so the field was set at the regulated temperature - a sequencing step that transfers to a cyclotron magnet.

Cited in: Cyclotron Magnet Design