irritating ground loop problem

[tempus]

Hey all;

My pedalboard is finally done! (I'll post photos shortly). I'm still working out some bugs, though - like this one:

Here's my setup: guitar (stereo out with piezo pickup on ring, magnetic on sleeve) => pedalboard in which is an XLR (pin 2 is magnetic in, pin 3 is a phantom powered piezo in with 9v going to a buffer in my guitar) => out to pan pedal which pans the between magnetic pickups (which go to fx on the pedalboard) and piezo which goes to a balancing transformer to the PA. Here's the circuit for the transformer part:

The pedalboard is powered with a wallwart, the chassis is stainless steel, and is only grounded to the DC ground from 1 point. Even the amp channel switching is not connected to ground to avoid ground loops. If I plug all 3 (amp, PA and pedalboard) into the same outlet, all is good. However, if the PA is plugged into a different outlet, I get hum as I sweep the pan pedal. IOW, it's basically hum free at either extreme - 100% piezo or 100% magnetic, but as I go from one to the other, there is an increase in hum until I'm at 100% again.
If I unplug the amp altogether, I get hum through the PA. If I connect the ground of the pedalboard to the ground of the PA (like the sleeve connections on the cable), everything works fine again, but there is a considerable increase in hum from the amp.

I kind of hoped I wouldn't have any ground loop issues with the balanced transformer out, but apparently I was mistaken. Any ideas on how to fix this mess?

[Paul Marossy]

I play a Parker Mojo Nitefly. I have built some splitter boxes for people that play "stereo" Parkers to replace those cheesy Y-cords that they give you with the guitar, which can cause ground loops since it connects the grounds of two different amps together, which in turn can cause a ground loop.

One of my variations is to have the 1/4" TRS cable from the guitar go to a stereo jack which splits the two input signals to a pair of jacks. The tip is the mag pickups which is connected to a regular 1/4" mono jack, for connecting to a standard electric guitar amp. For the piezo (ring), I have it connected to another 1/4" mono jack, and I have a switch that will allow you to also send that signal to a built-in XLR jack wired similar to the schematic found here: http://www.pic101.com/audio/anatomy_of_a_direct_box.htm

I have another switch, a ground lift that connects all of the grounds together or breaks that connection. This works for people 95% of the time. In some weird cases, isolating the grounds makes matters worse but a common ground doesn't work acceptably, either. I don't have an answer for those other than it might be a problem in the PA itself.

Anyway, is this similar to what you are trying to accomplish? Also, why do you have the resistor/cap combo on the output? On a direct box, the body of the XLR jack is also connected to ground on the PA side. This grounds the sleeve of the cable which is the cable shielding. If you don't have the cable shielding connected to ground, you essentially disable it and can get more noise, depending on how the grounding is done at the PA. Maybe that is part of your problem?

Also, a direct box typically uses a 10K/600 audio transformer. I think what you used is also contributing to the problem.

If you want to pan between peizo and magnetic, you'll also probably need to use a dual ganged pot wired opposite of eachother to keep both signals isolated.

[tempus]

One of my variations is to have the 1/4" TRS cable from the guitar go to a stereo jack which splits the two input signals to a pair of jacks. The tip is the mag pickups which is connected to a regular 1/4" mono jack, for connecting to a standard electric guitar amp. For the piezo (ring), I have it connected to another 1/4" mono jack

That's essentially what I've got happening.

Also, why do you have the resistor/cap combo on the output?

This is a damping circuit to keep the high end smooth by limiting unwanted ringing.

On a direct box, the body of the XLR jack is also connected to ground on the PA side. This grounds the sleeve of the cable which is the cable shielding. If you don't have the cable shielding connected to ground, you essentially disable it and can get more noise, depending on how the grounding is done at the PA. Maybe that is part of your problem?

I'll look into that. I forgot about that possibility. If I remember right, the cable shielding/sleeve is not connected to pin 1 to avoid ground loops, but I need to check with my meter to see what's connected to what. I do know that the XLR input sleeve is not connected to pin 1 or it would be connected to the chassis and we'd have ground loops again.

Also, a direct box typically uses a 10K/600 audio transformer. I think what you used is also contributing to the problem.

Hmm... I'm not sure why that would be - I'm pretty sure my PA specifies an input Z of 150 ohms, which is fairly typical for a mic input, and in any case, wouldn't a lower impedance be preferable to a higher one?

[Paul Marossy]

I see, a damper on the output. I guess that makes sense. I don't know that it's necessary, but I guess it can't hurt.

On the audio transformer, if that's what your PA says it wants to see, then use that. I think most passive direct boxes use a 10K/600 ohm transformer, though. I was thinking that perhaps feeding a PA that's looking for a 600 ohm load with a 150 ohm load would cause impedance mismatching, which might possibly contribute to more noise.

Hopefully you'll figure out what the problem is and it will work great for you after that.

[tempus]

Got it straightened out.

Just when you think you know about ground loops...

I realized that the schematic was slightly misleading, to myself as well as you guys. The damping circuit wasn't connected to the PA ground, it was connected to pin 1 of the XLR out. This didn't really affect anything as far as hum goes, but I poked around a bit more and found that it I connected the primary and secondary grounds to the signal ground, but not to pin 1, everything worked fine. So not isolating the grounds seems to ahve been the solution in this situation.

Go figure

Thanks

[Paul Marossy]

Cool man, glad you got it fixed. Yeah, sometimes NOT having something referenced to ground can cause just as much trouble.