[BradJudy]

I just finished hooking up my PPA in my new case. I tested it before installing it - worked fine. I tested it after installing it - worked fine. I tested it after connecting the input ground to the case - no sound at all. I detached the input ground from the case - works fine. I removed the output jack from the case to make sure the input and output grounds weren't connecting via the case (even though I isolated the output jack) - same behavior. Any ideas why connecting the input ground to the case would silence my PPA?

Thanks,

Brad

 

[Heady]

Just a wild guess. The DC power connector is not isolated from the chassis?

 

[BradJudy]

I should have mentioned, this is in a ParMetal case with a TEPS power supply. No part of the power supply is grounded to the case.

Here's a pic I took before I hooked up the wiring (and before the PPA board was mounted properly):

http://s94266093.onlinehome.us/head-fi/PPA_1.JPG

 

[morsel]

Use your DMM to make sure voltages are OK. Use your finger to make sure no components are overheating. Use your oscilloscope to measure the input signal at various stages through the amp to find out where you are losing signal. Are you tying input ground to the case at the input jacks? Are the input jacks shorted? Try detaching everything from the case except input ground, and putting them back one at a time until you find what the problem is. Make sure nothing on the underside of the pcbs are shorting to the case.

 

[BradJudy]

I had checked the voltages before attaching the input ground to the case (they were fine), but not afterward, so I will try that.

No components were overheating, but I will double-check that.

I am not tying the input groud at the jacks, but to a screw in the case (one of the PPA board support screws).

The input jacks are not shorted.

The PCBs are 1/4" off the bottom of the case, so nothing should be touching, but I will double-check that as well (in case a screw or somethign slid under the board).

I don't have an oscilloscope to measure the signal.

 

[morsel]

Quote:

I don't have an oscilloscope to measure the signal.

Doh! An oscilloscope should be mandatory equipment. You can use your DMM in AC mode to measure the input signal. I suggest using a 1k test tone. If you don't have a signal generator, buy or make a test tone CD and put the 1k tone track on infinite repeat, or make a 1k tone wav file and use your sound card as the source. Measure the signal level at the input jacks, after the volume control, after the opamps, and at the outputs. See if the volume control affects the signal level.

 

[BradJudy]

I had a surplus oscilloscope that allmost immediately died on me. Since I rarely do projects (2 CMOYs, a META42 and now a PPA/TEPS) it doesn't seem like a worthwhile investment (nor does a signal generator). My DMM (Fluke 87) provides what I need the vast majority of the time. It does have a frequency meter on it, so that (along with basic VAC) will help to troubleshoot the problem.

 

[tangent]

With a known-working amp, tracing a test signal through the amp will almost certainly tell you the problem. The signal is stopping at a very definite stage. The only question is, which one?

 

[BradJudy]

Okay - when the signal ground is not connected to the case, the potential between it and the output channels in nominal (couple of mV), but when the input ground is connected to the case, it jumps to ~3VDC!!

Thoughts?

 

[BradJudy]

I poked at this again. Just for giggles, I replaced my 637/627 opamps with 8610s for all three channels. Same behavior: with the signal ground not connected to the case, the DC offset is fine (~1.5mV), with it connected it jumps to 2.5-3.5V.

The TEPS appears fine - putting out an even 24V. I verified that the AC ground is not connected to the case at all.

I looked for stray screws or items that might be connecting the circuit to the case - didn't find anything.

The jacks appear to work fine - they are not shorted internally and they are isolated from the case where they are mounted.

The pot housing is grounded to the input ground, but is not touching the case.

So should I just try to trace the signal through different points in the circuit using the ACV DMM? If so, should my reference ground point always follow to a parallel point in the ground circuit? What should I be looking for?

 

[Voodoochile]

Measure between the PPA board's ground plane and the chassis. What do you find?

 

[BradJudy]

Quote:

Originally Posted by Voodoochile Measure between the PPA board's ground plane and the chassis. What do you find?

12.3VDC (naturally when they aren't connected) - hmmm....what would cause that?

 

[Voodoochile]

I think one side of your PSU is shorting to the chassis. Fine (not really!) with isolated input ground, not fine grounded. The thing is that your volume pot housing is also touching ground at the faceplate, though probably not very snugly. On an alps, it (pot housing) is common to the PPA ground plane.

 

[Voodoochile]

I suggest you check the nut near the heatsink on your PSU. It is very close to the sink typically, and the sink=output voltage. Output voltage to ground plane would equal approximate 1/2 voltage.

I either grind off the corner of the heatsink, or use a nylon nut for that one corner's bolt.

Of course your voltage leak might be coming from elsewhere.

 

[BradJudy]

Quote:

Originally Posted by Voodoochile I suggest you check the nut near the heatsink on your PSU. It is very close to the sink typically, and the sink=output voltage. Output voltage to ground plane would equal approximate 1/2 voltage. I either grind off the corner of the heatsink, or use a nylon nut for that one corner's bolt. Of course your voltage leak might be coming from elsewhere.

Give the man a cigar!!! I removed the nut and tested - nominal potential between the input ground and the case now. I figured the 12V meant something about the case being grounded to the power supply circuit, but didn't catch the nut to heatsink connection. I'll pick up a nylon nut for that position. I owe you man - if you see anything you want in my upcoming extra parts sale, it's yours.

I think this also means it's time to contribute to the Head-Fi fund - the countless help received here makes it one of the most productive audio forums out there.

BTW: My volume pot isn't actually grounded at the faceplate - it completely floats within the hole supported by the board mounts.