Thanks for the rabbit trail, Ethan... 
So naturally I haven't done much with the M-___ the past couple-three days, but my mind had been thinking here and there in the background about the whole ground-tie thing. I read that article a couple times that you linked a few posts ago, Ethan...I highly recommend it. It has got some history bits in there and it isn't just about grounding, but rather the whole subject of balanced and unbalanced cabling and connectivity and spells out the mess we're in very well. Well written. Before I read it I thought there was balanced and unbalanced; nothing more. I now realize there is right and wrong ways to do it, and that much of what you'll find in the marketplace doesn't do it right and then since there are differing ways to do it wrong the whole problem is compounded when you try and put all the stuff together. Balanced systems with balanced cabling done the right way (where the shield connects to chassis ground at both ends and does NOT act as a return path for audio signal) should never hum or buzz...but we've all dealt with hum and buzz in balanced line right? At least I know I have...so we get into band-aid fixes like ground lifts and isolation transformers to deal with a problem that shouldn't be there in the first place if the industry could agree and standardize on a proper convention...really, really interesting.
Read it.
So out of curiosity I looked at the mic in on the M-___...maybe I'm getting this all wrong but pin two is the shield and is connected to the audio ground, 0V(C), not the chassis ground. The M-500 mixers are the same way (looking at page 4-15 of the schematis). Not a deal breaker but interesting, and I'm not sure I understand how the ideal applies to that input...should there be a fourth conductor then as shield to chassis, and does pin 2 on the mic connect to it's chassis?
I'll probably read it again and get out the DMM and do some probing...
But another thing the article really kind of cleared the fog on is all the different kind of grounds:
- audio grounds: 0V reference voltages for +/- rails in the audio path like in opamps and such like the 0V(C) we have been discussing
- signal or noise grounds: similar to the first one but a separate run to keep noisy non-audio circuitry ground-path activity out of the audio chain like the 0V(D) we have been discussing
- grounds that go to the chassis
- grounds that go to earth
The last two are connection paths for the first two.
ANYWAY, what does all this have to do with the M-___ you ask? Well the question is still out there as to what to do with tying 0V(D) in with the rest of the grounding scheme. All of the channel modules only use 0V(C)...which begs the question "what happens to the non-audio circuit noise in the channel modules?" And this goes right along with a much earlier observation that the 6V rail also does not go to the channel modules, so all the LED's and such are powered by the 15V rails (different than the M-500 mixers where the 6V rail goes everywhere powering LED's)...maybe Teac hadn't figured that out in the M-___ and I'll be dealing with it (noise), or maybe they came up with some widgety way to iolate noisy circuits within the module. Dunno...but the 6V and 0V(D) rails obviously go to the master Control Module and the meter bridge, so back to the question "where do I tie 0V(D) into the ground scheme?" My current answer is found on pages 5 and 6 of the Rane article, particularly page 6. There it talks about the "star ground scheme", the official way of saying "all grounds need to tie in one place". I don't understand why Teac broke this simple rule of thumb with the M-500 mixers...everything but 0V(D) connect to the star in the PS-520, and 0V(D) is tied to 0V(C) in the Buss Master section...why did they do this? I'm sure there is a good reason, and I'm sure it has to do with avoiding noise entering the audio path...well, I see no evidence that they did the same thing in the M-___, and the M-___ is a very different creature than the M-500 mixers. So since it is a different creature, and since I cannot find where 0V(D) and 0V(C) have ever tied together in the mixer, I'm going to assume that the original PSU for the M-___ held a complete star ground scheme that included 0V(D), and this is a base ideal according to the Rane article. I'm assuming that the worst that can happen if I do this is that unwanted noise will enter the audio path and then I go back to the drawing board. But the star ground scheme is the ideal and I can't see that they did different in the M-___ so I'm going to start there unless anybody advises otherwise.
Since 0V(D) is isolated within the chassis of the PS-520, am I risking any damage by installing a jumper between terminal 7 of the PS PCB and the main ground trace?
I'm assuming not, and I will look closer and try to come to a comfortable answer myself, but input here would help me get there much faster and to a greater degree.
