I'm trying to determine the probable cause of a rectifier failure yesterday. I have 2 Mono EL34 push pull amps configured for a pair of 6AX4 for the rectifiers. I had my solid state plug-ins installed. While the amp was idling after running for a half hour or so I heard a pop, and one of the resistors in the plug-ins glowed bright and I shut the amps off. The plug-ins are each a 1N4007 followed by 100 ohms. The resistors are Welwyn WHS5 wirewound 100 ohm 5 watt: https://www.mouser.com/datasheet/2/414/TTRB_S_A0011516127_1-2565585.pdf.
Opened the amp up to check. The resistor that glowed is now open, the other measures in tolerance. Both 1N4007 test bad on the diode test (same voltage drop either direction). No signs of any other problems in the amp, B+ node to ground measures ok on resistance test. I tested the tubes, still like new with shorts test passing.
This morning I plugged in real 6AX4s in the affected amp, biased, fired it up, and it's working fine. So now I want to figure out if my plug-in rectifiers are under-spec'ed or if this was just a freak part failure. The rectifiers ran fine for a little over a year. The resistors are "high-surge" types, but maybe their 500V rating is insufficient? (440V B+). Or maybe a 1N4007 is not enough?
Opened the amp up to check. The resistor that glowed is now open, the other measures in tolerance. Both 1N4007 test bad on the diode test (same voltage drop either direction). No signs of any other problems in the amp, B+ node to ground measures ok on resistance test. I tested the tubes, still like new with shorts test passing.
This morning I plugged in real 6AX4s in the affected amp, biased, fired it up, and it's working fine. So now I want to figure out if my plug-in rectifiers are under-spec'ed or if this was just a freak part failure. The rectifiers ran fine for a little over a year. The resistors are "high-surge" types, but maybe their 500V rating is insufficient? (440V B+). Or maybe a 1N4007 is not enough?
Power transformer is 360v AC each side by spec (Dynaco PA 060), so that gives 440 - (-360) so 800V peak inverse, so if that 360V spec is RMS, then I'm about 100V over the 700V RMS spec on the 1N4007. So that checks out. So I guess I need 2 1N4007 in series, or a diode with higher PIV.
You could also exceed the peak rating when the amp starts from cold because the caps will charge immediately and the B+ will go up to 500 ++ volts until the outputs start conducting. You might be interested in watching that process on your meter, if not done so yet. See how close or over the working voltage rating for the caps that you get at start up.
Yes, although it survived power on many times and this failure was after running for a while. I measured the power on voltage as less than 500V when I built it, because I am using 500V rated caps.
