ZV5 Startup Troubles (Revised)

I am patiently building a ZV5 into an old Cutler-Hammer VFD case. The cool part is that it has a forced air cooling system that cools the heatsink (10” wide by 20” long by 6” deep).:hot: It works great!

First is that I cannot get the power supply to regulate. It is delivering aprox 4V less than the voltage that appears at the gates (which I can vary depending on the value of zener diodes I put in the circuit) but it seems to not to be regulating at a fixed voltage as the reg V sags whenever I add load to the output.

Regarless of this. I think I have bigger problems as I brought the audio circuit on-line this afternoon and it is definetly not working right. I have set the bias to aprx 560mv for each side and have my dc offset to within 20mv or so. When I present a signal at the input it sounds way off (very distorted with heavy echo etc.). And if I place a signal at only one channel it comes out of both speakers?

Well I guess I have some rework to do!. Anyone out there have a similar experience. If so I would love some advice.


Mad_K: The regulated voltage is about 3.5v less than the unregulated voltage and as the unregulated rails sag (when I apply load to the PS output) the regulated rails also sag (staying always 3.5v less than the unreg rails.

I have a single transformer (1000VA) and a single set (pair) of bridge rectifiers. The output of the bridge rectifiers feeds two seperate channels of cap, coil, cap, regulator, cap all per the schematic. IE I have mono supplys with the exception of the transformer and rectifiers.

JH: You may be right. I did use this calculation originally but now that you have me thinking about it again I may not have allowed sufficiently for the transformer regulation (IE it is spec'd as a 24vac secondaries but measures that at no-load).

I will try a lower value zener and expect that this may get the ps regulating properly. Still expect to have the same problem with the audio cct though!

Thanks for the ideas.


I envy at the 1K transformer . . . I always like to have it . . . big . . .
I prefer one big transformer covering both channels to two smaller . . .
because one big could cover both channels better way,
for example, in the music, there could be cases that the one side suddenly
requires 0.7 while the other needs only 0.3 or less . . .

I think that if the zener voltage is suitable value and the R13 and 14 are the sizes aimed for the current flow of about 2mA across them, you should have no problem in your psu. Your amp would work as wanted after the circuit having been carefully checked once again.

Looking forward to your impression of its sound. . .

Hope You get this sorted out, looks like Your problem is somewhere in the powersupply as this amp is very sensitive in that respect. The input is refered to the rails, not to ground, so any ac remaining will appear on the output, seems like the signal travels that way as You get output on both channels from one input. 24 Vac unloaded is a bit low, also bias should be at least 1A
per output device, preferably more.
Keep working with it, the amp is worth it!

Success! Well Sort of.

To all:

Thanks for all your replies. Based on the common suggestion that I have a problem in my power supply I tried the audio cct with a diferent power supply. Actually it was the original power supply that I was going to use but didn't because the noise was too high (1mv).

With this power supply the audio cct worked great, sounded pretty darn good to (free air through a couple of 3" TB's).

So I now know that you were all correct and the problem is in the power suppy. So, I have a few questions.
i) My regulator still does not appear to be regulating to a specific volatge. Vreg tracks Vunreg (less 3.5v) and the sag together under load). I understand that Vreg stability is desirable but I am assuming that this lack of stability is not the cause of why this PS causes my audio cct to mix one channel in with the other and generally sound like a peice of Racoon poop.
ii) Could the cause acutally just be that I have only one Transformer and set of bridge rectifiers feeding the two sets of caps/inductors/caps/regulators (one per each channel)?
iii) Or could the cause be proximity of the audio cct to the PS (aproximately 6 to 8". I did use some copper sheilding for a pair of the inductors that were particularily close 6"?
iv) Or could it be just a bad component somewhere?

Unfortunately I do not have a scope that can measure ccts above 20v so I am stuck in caveman troubleshooting land. Any advice is appreciated.

Last question, specifically for Mr. Pass. Based on my testing so far I think I can keep the Fets at 60c and the sinks at 48c with bias of 2.5 amps (fan tunnel works awesome). So, my question is whether there is a law of diminishing returns on the bias setting. Should I go for more. :devilr:

Sorry for the loooong pos and pls see link below for pictures of what used to be a VFD in a fish food plant!



I hardly understand it. . . I have likely used one transformer to cover both channels, without the mixture that you now experience. My claim is true as long as the ground keeps the potential zero at any condition and the signal doesn’t go to the other channel. If I were you, I would check the ground arrangement once again, among others. . .

jh, you rock!

jh, you were right!

I originally had the two PS channels connected to a bridge diode per Nelson's article. I guess when only using one transformer this is not a good idea. I connected all PS channel grounds and the signal grounds together with the AC ground and the chassis. The amp now works correctly and does not mix the channels.

Unfortunately I now have a ground loop. Fairly low volume hum with no imputs connected. With inputs connected it is quite loud. Weird part is that the hum mostly goes away at certain volume settings.

I will do some hunting around for this loop and also get some good speakers out to my garage for a proper listening test!

Thanks again jh!



Not at all . . .
One day you will help my question . . .
The hum is a devil . . .
Let's remember what Nelson Pass has said.
Particularly, avoid the input and output ground line passing through the psu caps . . .
And, let's remember the 0.7V lift between the earth and the ground . . . (or thermistor instead of the 0.7V lift) . . .

Good luck.
Again expecting to know your impression on the sound . . .