A 0-to-3000 volt power supply for biasing radiation detectors

This area is for discussions involving any fusion related radiation metrology issues. Neutrons are the key signature of fusion, but other radiations are of interest to the amateur fusioneer as well.
Post Reply
User avatar
Rich Feldman
Posts: 1351
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

A 0-to-3000 volt power supply for biasing radiation detectors

Post by Rich Feldman » Sun Feb 03, 2019 9:49 pm

Time to wrap up the bias supply and get to work on charge amplification for my BF3 tube.

The supply is made almost entirely from parts already on hand.
For starters (like the nail or stone from which to make soup), I had a pair of small variacs mounted on a bent strip of aluminum sheet. Removed one of them, to make room for the tiny NST featured early in NST Power thread. Added a proper removable cord connector, power switch, fuse, and indicator lamp.
The HV output connector is a pair of receptacle terminals, 3/4 inch apart like common dual banana plugs. Connected to NST secondary winding, a 0.01 uF plastic film capacitor, and one 30 kV 30 mA silicon diode. Some day I want to replace the latter with a 1X2B vacuum rectifier, which has a lower forward voltage drop up to 1 mA, and would make the whole design semiconductor-free.
The capacitor has no bleeder resistor when HV panel meter isn't connected -- a detail I was reminded of twice while changing wires for load test experiments. :-( Anyone want to guess what message is printed on the side of capacitor in picture?

HV circuit is floating, so it would be OK to connect either side to ground. Or both, without damage.

A panel meter is sort of built in, if we count the plywood as part of the enclosure. For this purpose, one kolovolt is defined to be 25 regular volts. Note that the variac came with tab terminals that extend beyond the edge of the original aluminum part.
Design has a 1750 ohm power resistor in series with NST primary winding, to soften the output & reduce risk of damaging detector tubes, without using a resistor on the HV side. e.g. 120 V from variac, 58 V dropped in resistor, 78 V across NST primary.

Here are some measured I-V characteristics at the output, with variac output adjusted to 120 volts AC.
DC current and voltage were measured at the same time, except for the red dots. Red currents were computed from the voltage and the separately measured load resistance. They hint at significant voltage dependence in plain old 2W carbon composition resistors -- see HV Divider thread for details. Most recent addition to the design is a bypass switch to unsoften the output, for when there's more time to investigate the red dot behavior.

The smoothing capacitor value was chosen to reduce HV ripple when load resistance is many tens of megohms. It doesn't do much about ripple with sub-megohm loads, as used for the higher current end of curves in chart.
All models are wrong; some models are useful. -- George Box

Post Reply