TL;DR: Building a DIY X-ray generator from scratch requires a high-voltage X-ray tube, a filament power supply, and a voltage multiplier circuit capable of delivering 50–80 kV DC. The most critical—and最容易踩坑的—components are the high-voltage ceramic disc capacitors in the Cockcroft-Walton multiplier. This guide walks you through the physics, the circuit design, real test results, and the safety protocols you absolutely cannot skip.

⚠ WARNING / DISCLAIMER

This article is published for educational and theoretical reference only. X-ray generators produce ionizing radiation, which poses serious, cumulative health risks including radiation burns, DNA damage, and increased cancer risk. High-voltage circuits (>1 kV) carry lethal electrocution and arc flash hazards. The author fully dismantled the system immediately after completing this photographic project.

Do NOT attempt to replicate this project without professional radiation safety training, calibrated dosimetry equipment (Geiger-Müller counter or scintillation detector), proper lead shielding, and a remote-operated setup. The author assumes no liability for injuries, property damage, or legal consequences resulting from imitation of any technique described herein. Check your local regulations—unlicensed X-ray device operation may violate federal and state radiation control laws.

Background — How This Project Came Together

A friend of mine managed to salvage a surplus medical X-ray tube and shipped it over to my bench for free. As a hardware hacker, you simply don't let a perfectly good X-ray tube sit on the shelf gathering dust. So the mission was clear: build a driver from scratch and see if we could get some photons flying.

By "light it up," I don't mean making it glow like a neon sign—I mean getting it to emit actual X-rays. And to do that, we need to talk physics first.

How X-rays Actually Work

X-rays were discovered by Wilhelm Röntgen in 1895 while he was poking around with cathode ray tubes. He noticed an invisible ray that could expose photographic film and pass through solid objects—and he named it "X" because he had no idea what it was. (Peak science energy: "I don't know what this is, but it's amazing.")

Modern physics tells us X-rays are produced by two mechanisms when high-speed electrons slam into a target material:

Bremsstrahlung (Braking Radiation): When high-speed electrons violently decelerate near an atomic nucleus, that sudden loss of kinetic energy has to go somewhere—boom, it's released as continuous-spectrum X-ray photons. Think of it as an electron "slamming the brakes" and spewing electromagnetic radiation as a byproduct.

Characteristic Radiation: When an incoming electron knocks out an inner-shell electron, an outer-shell electron drops down to fill the vacancy—and emits a photon at a frequency unique to that element. These show up as sharp peaks in the spectrum and are useful for identifying materials.

The X-ray Tube — Anatomy of the Beast

Any X-ray tube needs three things:

(1) Electron source — A heated tungsten filament (yes, same as in light bulbs) that boils off electrons via thermionic emission.
(2) High-speed electron beam — Created by applying tens of kV between cathode and anode, inside a hard vacuum to minimize collisions with gas molecules.
(3) Target material — A tungsten anode with extreme density (19.3 g/cm³) and melting point (3,422°C). The electrons slam into this target, decelerate violently, and produce X-rays.

Thermal reality check: Less than 1% of the electrical energy becomes X-rays. The other 99% turns into pure heat at the target. At 100 kV and 2 mA, that's ~200 W concentrated on a spot the size of a pinhead—comparable to a laser cutter. Without oil cooling, your tube turns into an expensive paperweight in seconds.

My tube is a small-power unit with oil-immersed cooling. For short DIY exposures at low current (1–2 mA), this is perfectly adequate. Running at higher currents without active cooling is a fast track to melting your anode.

Soft vs. Hard X-rays & The Physics You Need

X-ray energy is measured in keV, and in a tube, roughly 1 kV of accelerating voltage = 1 keV of X-ray photon energy.

Key distinction: Voltage (kV) controls penetration power (photon energy / wavelength). Current (mA) controls brightness (photon count). Raising kV makes X-rays that punch through denser material. Raising mA gives you more photons for shorter exposures—useful when you want to minimize radiation dose per image.

Below 25 keV: soft X-rays — easily absorbed by glass and tissue. These need special beryllium windows to escape. Above 25 keV: hard X-rays — the kind used for imaging and industrial inspection.

Lead shielding myth: Lead doesn't magically block all X-rays—it attenuates them via its high atomic mass and thickness. A thin sheet of lead foil will be penetrated by high-energy X-rays. For serious shielding, you need thickness (>1 mm for hard X-rays) or higher-Z materials like depleted uranium (yes, radioactive material shielding radiation—physics is weird).

The Build — From Zero to X-rays

Attempt #1: Flyback Transformer Hack (Failed)

My first instinct was to avoid capacitor voltage multipliers entirely (more on why later). Instead, I wired four black-and-white TV flyback transformers in series with combined EE cores—pure junk-pile engineering. I encapsulated the whole assembly in paraffin wax, pulled a vacuum to eliminate bubbles, and fired it up.

Result: ~50 kV output, followed by rapid breakdown. Paraffin shrinks during solidification, creating micro-cracks that kill insulation integrity at these voltages. Dead end. Back to the voltage multiplier approach.

Attempt #2: Cockcroft-Walton Voltage Multiplier

Transformer: A custom UY20-frame step-up transformer rated for 15 kV / 20 mA output (300 W). For a 6-stage multiplier targeting 75 kV, this is the minimum starting point. Total cost: ~$15 USD from a specialty manufacturer.

Capacitors: 35 kV rated, 2200 pF high-voltage ceramic disc capacitors. Rule of thumb: each capacitor must be rated for at least 2× the peak input voltage per stage. These are not designed for high-current discharge—treat them like static capacitors. Avoid direct spark-gap testing after assembly; it can damage them.

Diodes: Salvaged 20 kV / 0.5 A silicon rectifiers, two in series per stage (giving ~40 kV breakdown per stage). For a clean build, source 30 kV / 20 mA silicon diode stacks—widely available on AliExpress and eBay.

Insulation: First lesson learned: paraffin wax is useless above 30 kV. The entire multiplier assembly now sits in silicone oil (mineral oil or transformer oil also work). Never use regular lubricating oil—it's not rated for dielectric stress.

Wiring: Multiplier output runs through 80 kV-rated high-voltage silicone wire. For the filament supply, I skipped fancy DC-DC converters (afraid of HV feedback killing them) and went with a single lithium cell plus a length of nichrome heating wire as a current limiter. Simple, reliable, and nearly indestructible.

Test result: 24 V input to the primary transformer easily arcs across a 6 cm air gap. That's roughly 180 kV peak (air breakdown ≈ 30 kV/cm at STP)—confirming the multiplier is working hard.

Imaging Setup — Intensifying Screens & Polaroid Film

For actual X-ray imaging, I used rare-earth phosphor intensifying screens (green-emitting, ~10× brighter than the blue screens I started with). The film goes in a light-tight cassette pressed against the screen—over 90% of the exposure comes from the screen's fluorescence, not direct X-ray impact.

Polaroid hack: For the final images, I used a broken Polaroid camera's film ejection mechanism (scavenged off eBay for $15). The workflow: extract Polaroid film, expose it in a standard X-ray cassette with intensifying screens, then reinsert it into the Polaroid for instant development. Surprisingly effective.

Results: Internal components of a DC-DC boost converter, iPhone 4 internals, keychain flashlight structure, and yes—my own finger bones. All captured on Polaroid instant film with the green phosphor screen.

Radiation Safety — The Part You Cannot Skip

During testing, I monitored radiation levels continuously with a Geiger-Müller counter and maintained a minimum 5-meter standoff distance with remote triggering—following the ALARA principle (As Low As Reasonably Achievable).

The inverse-square law is your best friend: Double your distance from the source, and radiation exposure drops to one-quarter. This is why distance is more effective than most shielding.

Even facing away from the tube window, X-rays scattered off the building's steel rebar caused the Geiger counter to peg at full scale. This fully exposed, unshielded setup is not safe for anything beyond a brief demonstration. A proper enclosure with lead-lined walls is mandatory for any sustained operation.

The Critical Component — High-Voltage Capacitors

The heart of any voltage multiplier is the capacitor bank—and this is where most DIY builds fail. You need capacitors that can:

• Withstand extreme DC voltages (30–50 kV per stage)
• Maintain stable capacitance under repeated charge/discharge cycles
• Handle the high dV/dt transients of a Cockcroft-Walton circuit
• Resist voltage coefficient degradation (capacitance dropping under DC bias)

High-voltage ceramic disc capacitors (N4700 or equivalent formulations) are the standard choice for this application. They offer excellent voltage withstand, low dissipation factor, and compact form factor for tight multiplier layouts.

At HVC, we manufacture high-voltage ceramic disc capacitors specifically designed for demanding applications like X-ray power supplies, particle accelerators, and high-energy physics. Our N4700-series capacitors are rated from 1 kV to 50 kV+ and trusted by Fortune 500 companies worldwide. If you're building a similar project, choosing the right capacitor is the difference between a reliable system and a catastrophic failure. Browse our HV Ceramic Capacitor catalog or contact our engineering team for application support.

Builder's Cheat Sheet — Quick Reference

Parameter	What It Controls	Practical Range	Watch Out
Voltage (kV)	Penetration power (photon energy)	50–100 kV for imaging	Higher = harder to shield. 100 kV easily penetrates thin aluminum.
Current (mA)	Beam intensity (photon count)	1–5 mA typical	Higher = more heat at target. Keep duty cycle low without oil cooling.
Soft X-rays	Below 25 keV	Blocked by glass	Need beryllium window to exit tube. Mammography uses these.
Hard X-rays	Above 25 keV	Standard imaging range	Penetrates glass. Requires proper shielding and distance protocols.
Duty Cycle	Tube survival	0.5s on / 2min off (air cooled)	Without oil bath, continuous operation = dead tube in seconds.

Pro-Tips & Common Pitfalls

Don't use regular lubricating oil for insulation. It's not rated for dielectric stress at these voltages. Use mineral oil (food-grade, also used in PC oil-cooling builds), transformer oil (mineral oil with additives), or silicone oil. The oil must be clean, dry, and free of air bubbles—moisture is the enemy of high-voltage insulation.

Tungsten target thermal management is everything. The target absorbs 99%+ of input power as heat. Without active oil cooling, keep your duty cycle extremely low: 0.5 seconds ON, then 2+ minutes OFF to dissipate heat. If you see the target glowing red, you're already too late.

Strip copper foil around all solder joints on your high-voltage board. At >30 kV, surface tracking across PCB copper is a real failure mode. A 5 mm clearance around every joint is a good starting point.

Corona discharge is normal—but manage it. At high voltage, you'll see a purplish glow (corona) around exposed conductors. It's not dangerous per se, but it generates ozone and can degrade insulation over time. Encapsulate everything in oil.

Going Further — Ideas for Version 2.0

• Digital Radiography (DR) upgrade: Replace Polaroid film with a medical-grade CCD sensor behind a scintillator screen. Old flat-panel detectors from decommissioned DR systems occasionally appear on eBay/surplus sites and can be adapted for real-time X-ray imaging.

• Marx Generator: For higher peak voltages with faster rise times, consider a Marx generator (parallel charge, series discharge) instead of the Cockcroft-Walton multiplier. Better suited for pulsed X-ray sources.

• Automated safety interlock: Add a door/window switch, radiation monitor, and automatic HV shutdown to create a proper safety enclosure. This is not optional for anything beyond a one-time bench test.

• Spectrum characterization: Use a semiconductor detector (Si-PIN or CdTe) to measure the actual X-ray spectrum output of your tube at various voltages. This data is invaluable for optimizing imaging parameters.

Summary

Building a DIY X-ray generator is a fascinating deep-dive into high-voltage engineering, vacuum physics, and radiation science. The key takeaways:

• X-ray physics: Bremsstrahlung + characteristic radiation. Voltage = penetration, current = brightness.
• Circuit design: Cockcroft-Walton multiplier with HV ceramic disc capacitors is the reliable approach.
• Thermal management: 99% of input energy becomes heat. Oil cooling or very low duty cycle is mandatory.
• Imaging: Rare-earth intensifying screens + Polaroid film is a viable low-cost approach.
• Safety: ALARA principle. Distance is your best shield. Monitor with a Geiger counter. Never skip the lead enclosure.

Final Warning: This project involves ionizing radiation and lethal voltages. The author fully dismantled the system after this photographic session. Do not attempt without proper training, shielding, dosimetry, and remote operation. Always maintain ALARA protocols. Check local regulations before building—unlicensed X-ray devices may violate radiation control laws in your jurisdiction.