This project started with a simple goal: build a compact, reasonably low-noise bias supply for a surplus PMT for spectroscopy experiments. Target output was around 1.4 kV at low current with acceptable ripple and minimal EMI coupling into the front-end.
Design goals
- Output: ~1.4 kV at a few hundred microamps
- Ripple low enough for spectroscopy use
- Small footprint so the system stays portable
- Easy to service without exposing high-voltage nodes
First prototype lessons
The first board worked — in the “electricity came out” sense — but it had measurable issues. Noise coupling into the amplifier, sloppy return currents, no attempt to segregate high dV/dt nodes, and cables acting like miniature loop antennas.
- Poor placement of the switching node
- Ground plane return paths were guessing their way around the board
- Unshielded leads happily radiated broadband garbage
Next steps
The next revision will include:
- A physically isolated switching node
- Proper shield can or copper tape enclosure
- Additional RC and LC filtering
- Measuring ripple directly at the PMT anode in a dark box
Once the supply is cleaned up, I’ll pair it with a proper front-end and begin characterizing spectra from various sources.