two amos with polarity switches tripping a gfci

[georage]

hi there. this is sort of a rhetorical question (not sure if that is allowed), because i will not be able to take these amps apart and confirm anything. so...
i installed a very nice gfci receptacle at a guitar store to serve the "amp island" display. all the amps in the store get their power from this one power outlet. i made the outlet gfci protected because of the extra risk of use by the general public.
If you dont know, a gfci device monitors the current between hot and neutral conductors of an ac mains receptacle, if they are different the gfci assumes a person is being shocked or a ground fault is accuring and shuts off.

the issue; when a music man amp and a mesa boogie amp are plugged in at the same time, the gfci trips, shutting down the electric because a ground fault has accured.

the music man has a 3 prong cord and a 2-way polarity switch. the mesa boogie has a 3 prong cord and a 3 position (+,-,ground lift) switch. when the two switches are positioned in one way the gfci trips, if they are positioned in another way, they work fine and dont trip the gfci. (does that make sense)

im just wondering if anyone has an idea of why this is happening? my guess is that the music man amp is mis-wired or malfunctioning. can anyone think of a way that 2 amps(plugged into the same gfci) can trip a gfci while being properly wired?

[georage]

ps cant change the post title now xD

[Jon Snell]

As the amplifiers have a three pole earthed mains lead and I assume the mains supply has been tested for correct wiring and loads, the capacitors that go between the 'polarity' selector and either pole of the mains to the chassis will present a current flow and trip the protection device.
Remove the capacitors after ensuring the amplifiers are wired correctly. They should not be fitted to three cored earthed (grounded) equipment.

The 'polarity'switch was originally intended for two pole mains cable fed electrical equipmet and brought the chassis of the equipment to the desired phase of the mains, hopefully reducing the hum by allowing the mains borne interference to return to the mains. It did kill or seriously hurt quite a few musicians.

Imagine, the amplifier connected to the live wire through a faulty capacitor and the PA has a properly grouned chassis;
The guitar is at mains voltage and the microphone is at ground.
The outcome is obvious.

Fitting the GFCI is a good thing (RCD if done by a professional in the UK) as it does stop electric shocks from faulty or non compliant equipment.
110volts in the USA is not as lethal as 240volts elsewhere!

[Tom Phillips]

I agree with Jon.

Since the Mesa and Musicman amps have three wire power cords AND polarity (Ground) switches I'm thinking that the three wire power cords were added after the amps were originally built. If so, it is highly likely that the conversions were done incorrectly. I say this because almost all conversions that come into my shop were done incorrectly with respect to the connection of the hot and neutral lines.

[g1]

There was a time when amps had 3 prong as well as polarity switches with caps. (SF Twins etc.)
And some modern manufactured amps (and other equipment) use line to chassis caps, although class Y are required now. Actually quite common to see for EMI/RFI filtering..
Polarity switches are pretty much gone now though.

[georage]

unfortunately i wil never be able to take both amps apart and checkbtheir wiring. thanks for the input, though. as an electrician, personal safety is my main concern(followed by fire safety). a gfci works by monitoring the balance between mains hot and neutral currents, and tripping if the balance is not equal. If a hot or neutral current goes to the third pin(equipment ground), the gfci will trip.

It is interesting that either amp would not trip the gfci in any position of the polarity switch, but a certain combination of both amps would trip it instantly. Ive never experimented with capacitors in a gfci protected circuit, it might be normal or it might be a fault in the amps, idk. im happy my gcfi tripped, but i still wonder why. i think my mains circuit would still be valuable in a live situation where a soundcheck could reveal any faults even if it is annoying. i like the simple test of touching the guitar strings to the microphone before a gig. simple and safe.

[Greg Robinson]

Your GFCI trips at 10mA right? Any amp that has a "death cap" - common in US amps right up until the early 90's to cater for legacy ungrounded sockets - though these amps were fitted with 3 pin cords, they were still expected to function with ungrounded outlets by AC grounding the chassis. Depending on the setting of the polarity switch/correct polarity of the outlet, this cap will be between chassis and live, or chassis and neutral. If between live and chassis, this will then present a leakage path to ground if connected to a grounded socket. 0.1uF is a fairly common value for death caps, and has an impedance of 26k5 at 60Hz, allowing almost 5ma of leakage current (certainly possible when you consider the common 20% tolerance of caps).
This is cumulative for all appliances on the same GFCI circuit, so quite easy to surpass that 10mA threshold with two or more amps (there's also small leakage for any appliance/cabling due to capacitive effects between conductors - this is unavoidable, but also orders of magnitude smaller than the death caps.).

With the polarity/ground switch set appropriately (death cap to neutral or lifted completely), this leakage current will reduce/be eliminated. There'll still be a small leakage even with the cap neutral as impedances are non-zero (offset voltage at socket between neutral and ground even though they're bonded at the panel).

Proper fix is to remove the deathcaps as they're a dangerous legacy. If you're not in a position to do so, the practice here in Aus/UK/etc would be to fail it per "test and tag" - essentially label it as "withdrawn from service". I know this practice isn't so common in the US (otherside of internal company policy), but may be an option if this is still concerning you.

[georage]

It looks like the eaton brand gfci trips at 5ma. i did not realize a death cap would have a resistance even when properly working. Would that resistance be measurable with an ohm meter? I thought a capacitor should read no continuity when working properly. Come to think of it, i do see capacitors charging with my analog ohm meter...the needle jumps up showing resistance and then slowly subsiding to infinite ohms. thats an old fashion capacitor test. is that what you mean, Greg, when you said a death cap has a resistance, or is it a constant resistance?

Ive often wondered what the problem with a death cap is, besides the threat of a short or leaky cap. most modern power supplies i see have x and y safety caps. y safety caps look just like death caps to me, but they dont seem to trip GFCI's.

[The Dude]

He did not say resistance, he said impedance which is an entirely different thing. No, you can not measure impedance with a DVM. I've run into this exact problem with two rack mount power amps and a mixer on the same GFCI. Any one unit showed no problems by itself, but when all were used, the GFCI tripped. It was necessary to lower the cap values of some RFI filter caps on the AC input. The caps were just large enough and conducted just enough to trip the circuit when multiple amps were connected. Installing lower value caps solved the problem, but still provided adequate RFI filtering. Since no schematic was provided here, I can't say for sure that is your problem, but it sure sounds like it could be a death cap or other cap issue given that the problem does not exist with one amp at a time.

Edit: I found my original thread describing a similar problem, if it helps.

https://music-electronics-forum.com/...2000-trips-gfi