"Yellow sheet" diode mode, WZ68 copper cap, and start-up arcing

The backstory: I got a set of KT 120 tubes back in October for my ST120. I had previously done the "yellow sheet" modification when I was using a GZ34 rectifier. I then transitioned to the WZ68 copper cap. Lastly, I built a TDR module set for about a 15 second HV delay.

Recently, there was a rather nasty snap and flash I could see and hear around V7 when the TDR sent high voltage to the rectifier. The amp sounded fine after the snap. For some strange reason, when I replaced the KT 120's with my older Gold Lion reissue KT88's there was no start-up arcing but this could simply be a coincidence. I carefully checked everything visually under the hood and could see no obvious issues. I did, however reflow all the tube socket connections. I checked the tube socket resistors in circuit and found the bias resistors for V2 and V3 were reading a little over 14 ohms, those for V6 and V7 were very close to 10 ohms. The 1 K ohm resistors were all fine right at 1 K. I replaced the bias resistor on V2 with a resistor reading 10 ohms out of circuit and it read 14 ohms in circuit.

Now to my questions. In testing the voltages, with diodes connecting rectifier pins 6 and 7 and 4 and 5, I got the correct 430 VAC before the diodes, but only about 225 VAC after the diodes. I didn't expect this kind of voltage drop. 1) what caused the big voltage drop, just the AC to DC conversion or something else? 2) since I'm using a SS rectifier now, are the "yellow sheet" diodes even necessary or useful?

I removed the "yellow sheet" diodes and moved the AC high voltage transformer leads back to pins 4 and 6. Now with the KT 120's back in the amp, I haven't experienced the start-up arcing and flash over 5 or 6 start-ups. Maybe another coincidence, but perhaps all my tinkering did deal with the underlying problems. I haven't checked any other voltages but perhaps I should. Any insights to this situation would be appreciated!

Bill

Bill,

By using the yellow sheet mod with a solid state rectifer you are effectively just having 2 diodes in series from each HV winding. Not a big deal and will not cause any issues.

https://photos.app.goo.gl/GFSmMbqzwgc8GGko8

The reason you got a strange voltage reading is that your meter is reading the half wave rectified and unsmoothed voltage at the junction of the two diodes in series.

Steve

To clarify a bit:

a) A (typical) silicon diode gives about 1V drop. So, two in series will drop about 2V.
b) You were not measuring AC after the diodes, you were measuring chopped DC.
c) A meter on AC will measure chopped DC far less efficiently than AC, so try setting your meter to DC and remeasure.
d) THEORY: With an input voltage of 430 VAC and a pair of diodes in series, you should get ~300 VDC. The drop to 225 was because you were measuring the chop, not true AC.
e) Math is: 0.707 x Input AC = output DC for a single diode or diodes in series. 1.414 x Input AC = output DC for a bridge or full-wave rectifier. Both less associated drops.

NOTE: Some meters are smart enough to distinguish between chopped DC and AC, some are not. There is no shame in the one choice, and although the advantages are significant with the other, the expense may overcome the advantage. As long as you understand the results.

https://www.youtube.com/watch?v=LZEie3nAo1E

Gives examples of "True RMS, True RMS AC/DC, and Averaging meters.

another reason why I refuse to use an "autorangeing" meter which also is usually auto-function
I prefer to know I'm measuring AC or DC and on a 600v scale rather than some unknown scale

tubes4hifi wrote:another reason why I refuse to use an "autorangeing" meter which also is usually auto-function
I prefer to know I'm measuring AC or DC and on a 600v scale rather than some unknown scale

To be absolutely honest, here, I am not sure how to respond without being snarky. I will try.

I keep a Fluke True RMS/AC/DC meter. It is also autoranging. This allows me to:

a) Measure AC over DC where the AC component may be a few volts, or less and the DC component up to 1,000 Volts. This is very useful when diagnosing for hum loops, bad power-supplies, weak/failed diodes and similar disorders. And it helps that the meter is smart enough to know the difference between AC and chopped DC. And, did I mention DC offset?

b) Not have to switch scales when measuring at multiple test points. A meter in the 600V scale is not ideal for measuring 0.5 V or so. Nor will it blow up were I to forget to change back to the higher scale.

Good meters (tools) are not cheap.
Cheap meters (tools) are not good.
I would rather have multiple meters (tools) that do a few things very well than one that does many things not-so-well.
When dealing with tube equipment, and that which also carries lethal voltages that are immediately accessible, we want tools that are as forgiving (user-friendly) as possible, whilst remaining flexible and resilient. The range-of-skills in this group span from "nothing sharper than a rubber spoon" to "able to disassemble and reassemble a 787-8 blindfolded". A good tool is accessible to that entire range.

But, then, God invented the eraser because none of us are perfect!