Construction Classics

Electrical Design of Electrostatic Deflectors for Sector-Focused Cyclotrons

Why this document can be hosted

UCRL-10654 is a Lawrence Radiation Laboratory report produced under AEC contract W-7405-eng-48, which appears on the title page of this scan. AEC contractor reports are treated as US Government works and carry no copyright; pre-1978 distribution without a copyright notice is an independent public-domain basis. The scan carries the standard US Government sponsorship disclaimer and the University of California legal notice, and no copyright notice, publisher imprint, or journal-reprint marking appears anywhere in the document. Verified against the scan and the OSTI record; determination made August 2026.

Reading guide

Beam extraction is the thinnest subject in the amateur literature, and this is the report that fills it. It is not a theory paper: two Berkeley engineers describe what they measured while building the electrostatic deflector that pulled 50 MeV protons out of the 88-Inch Cyclotron, including the materials that failed and why. Page references below are PDF pages; the report’s own printed numbers run four behind, with disclaimer and legal-notice sheets bound in front.

The introduction (p. 6) sets the problem in one sentence worth memorising: “the key to deflector design is the art of tailoring the electrodes to the geometry of the beam.” The numbers that follow are the ones a builder needs for scale — the 88-Inch required a peak gradient of 150 kV per cm, and at extraction radius the useful beam stood only about 0.25 inch tall. That ratio, not the absolute voltage, is what makes extraction hard.

The high-voltage sections (pp. 8–11) are the transferable core. They set out the VE relationship — for equal probability of sparking with a given material, the product of gap voltage and cathode gradient is a constant — and then rank electrode materials by spark damage measured rather than assumed: stainless steel, inconel, molybdenum, K-monel, titanium and nickel best; copper, tantalum and aluminium intermediate; silver worst. Carbon resists spark damage well but will not stay baked out, and after the voltage is off even a few minutes the whole bake-in has to start again. One detail no textbook mentions and every builder should know: the spark dust from K-monel and nickel is magnetic. Figure 1 (p. 9) plots DC breakdown voltage against gap for seventeen materials, from invar and 316 stainless to lead, with test conditions stated.

The test-model deflector (pp. 14 onward) is the experimental heart, and the conclusion (p. 33) is where it pays off. The model reached a VE number of 2.25 × 10⁴ (kV)²/cm, but the authors recommend designing to 1.5 × 10⁴ to leave margin for day-to-day operation — a stated derating factor, which is exactly the kind of guidance amateur extraction attempts lack. The septum comparison on the same page is the other keeper: a tungsten septum did not reduce the VE number in the presence of beam, while a carbon septum held about three-quarters as much and only with no beam present — under an intense beam, evaporated carbon contaminates the high-voltage electrode badly enough to drop the deflector to a quarter of its normal holding voltage, recoverable only by venting to air and cleaning with solvents. Carbon still earns its place for short half-lives and low activation, so the choice is a real trade rather than a ranking.

Also worth the detour: ion scrubbing to strip surface oxides typically cuts deflector dark current by about a factor of five (p. 33). The machine is far larger than anything built at home, but the failure modes are scale-free — sparking, septum erosion, contamination of the high-voltage surface — and this is the only hosted document that quantifies them.

Design rules extracted from this document

14 Design Guide rules cite UCRL-10654. It was selected to close the extraction gap and it did: these are the measured numbers behind electrostatic deflector design, not estimates. Representative examples:

• dg-742 — size deflector gaps by the VE relationship
• dg-743 — derate to VE = 1.5e4 (kV)²/cm for daily operation though 2.25e4 was held in tests
• dg-744 — electrode materials ranked by measured spark damage in a magnetic field
• dg-745 — each electrode material has a critical field, 4–15 kG, above which damage is severe
• dg-746 — spark energy has an optimum, not a minimum
• dg-747 — budget conditioning at roughly 30 sparks per cm² of high-voltage surface
• dg-750 — support the high-voltage electrode at both ends — a cantilevered bar self-oscillates

Used by the beam extraction deep dive. For the supply that feeds these electrodes, see UCRL-10655.