Fusor Forums

While using a TC 174 the ground connection was lost to the BF3 tube (due to the detector separating from the housing and providing an intermittent connection) and presumably that's what damaged the preamp. I replaced both JFETs and ran test pulses through the preamp and everything appeared to be working correctly. I then repaired the BF3 tube with copper tape/conductive adhesive and verified that the ground and hv connection were good and isolated (with ohm meter). Unfortunately, once I reconnected the BF3 tube to the preamp and very slowly applied bias voltage, the preamp failed (I think immediately). Is it possible that the high voltage ac coupling cap was damaged the first time the preamp failed? It measures open with an ohm meter but possibly it lost some isolation capability and now passes HV? Thoughts?

Edit: I have now replaced the HV ac coupling cap, but I'm going to wait to power-up with the detector until someone more knowledgeable than me responds.

ceramic caps have a nasty habit of being a bit like zeners when damaged
they measure perfect on a cap meter but as voltage goes up they suddenly conduct
I have had many glassman supply multipl,ier stages with this exact fault usually somewhere in the middle of the multiplier.
the only way is to megger the cap at near its max voltage

John,

Good to know, but the ac coupling cap in the Tc 174 is a film capacitor.

Update:

I decided to power the BF3 tube through a 22Meg resistor and ac couple a scope probe using a 270p capacitor on a suspicion that the detector may be the cause of the preamp failure. At about 1100V (the manufacturers lower end of the recommended operating voltage range) a constant 500 mV p-p noise appears (~1kHz) and then disappears once the bias voltage is raised to 1200V. If I decrease the voltage back to 1100V the noise reappears with a greater amplitude. Is this a result of some unstable leakage current due to the tube anode being damaged? The noise while making detection problematic the p-p amplitude doesn't seem enough to damage the JFET.

Edit:
The scope has a lower input impedance than a JFET so the peak voltage of the noise when using the preamp may be higher.

Can you monitor the current to the tube as the voltage is raised and lowered. Sounds like some sort of voltage sensitive oscillation is set up in your circuit with the tube being the vacuum tube/neon lamp equivalent. Tough to noodle out.

Richard Hull

I returned the tube to LND for evaluation.

I'm currently using a SNM-11 corona tube and PAD 814 for detection and it seems to be working really well.