Last night I came to at least a working resolution of a really weird problem. I'm hoping others can tell me what is really going on here, and what best to do about it. I am not an electrical engineer or electrician or anything like that, I just learned a lot from my dad and friends and trial and error.

This has to do with my MIDI synth, which is a headless Linux box, often used with three Yoshimis running simultaneously. Input is just MIDI, either through USB direct from keyboards or through USB MIDI adapters. The problem is other really odd high-pitched non-static signals which are contaminating the audio, when plugged into any of three different PAs.

The contamination has occurred no matter whether I use motherboard 1/8" jacks, FrontX RCA jacks wired into the motherboard, custom-built Switchcraft 1/4" TS jacks wired into the motherboard, a Prosonus USB, or a Behringer FCA202. It occurs to varying degrees, not predictably, with all of them. AC power ground lifts sometimes have helped, sometimes not, but not predictably: plugging into the same PC, one week the extra signals will be there, the next week about 75% less, et cetera. Same with different MIDI methods. Once I knocked out 90% plus by going from USB direct to MIDI through the Presonus, and then two weeks after that it went up again (still using the Presonus) to about half what it had been. I have tried AC ground lifts on amp, keyboard, and the box. Sometimes the signal goes away when there is nothing connected via USB; sometimes not. Good AC grounding has been explicitly made excellent and tested very thoroughly by at least two professional musicians at one of these locations.

Perhaps the oddest thing I have noticed about the extra signal, sometimes, is when it comes. I usually connect the synth box through a Roland volume pedal to PA. If the pedal is entirely absent, it always comes. But if the pedals are there, about half of the time, the extra signal comes in most when the pedal is in the lowest/zero volume position, and least/zero when the pedal is in the highest/maximum volume position. The other half of the time, the signal is always there and always the same volume, no matter what the position of the pedal!

About three days ago, reviewing for the umpteenth time the whole memory-history of this project, I remembered one thing I did a long time ago which seemed to knock out the extra signal altogether (back then, it was a 60-cycle hum): I used a Radio Shack inline-audio ground lift device, a 2" black barrel with two short double-RCA-male cables. I wore out several sets of adapters on that thing which is why I abandoned it. But it had worked more positively than anything else I tried. And I found other reports on the WWW of similar setups not necessarily involving 60-cycle in the problem signal. So I investigated the state of the art of inline audio ground lifts, found a whole lot more easily findable options than there had been then (might have been ten years ago), found that Behringer among others is making one using 1/4" TS/TRS, ordered the Behringer, tried it last night, and voila, lovely gorgeous silence until the Great Key-Pressing!!!

So. I would like to rigorize this outfit a bit more; right now I have to play using the box plugged via Firewire into the FCA202, thence via TS into the ground-lift, and then to the PA via another TS. Here are a few questions.

2. What is the real advantage of TRS vs. TS? Some of the audio ground-lifts advertise TS/TRS conversion and/or problem resolution; what is the deal??? Does TRS handle longer runs better? Whenever I have the case plate reengineered, should I have it done with TRS?

3. The 96 kHz audio available via Firewire is clearly helpful in the upper registers. Fitting it along with a 1/4" double inline audio ground lift device into my box, is somewhat less than appealing, especially because the metal cases of both the Fireware and the ground lift devices, will be in contact with PC case metal, possibly resulting in more contamination. What to do ???!???