Fender Bassman Compact trips GFCI Breaker...sometimes

[tdlunsfo]

I had a new shop built at the house over the winter and apparently the new code is to put GFCI breakers in the panel now. Well, I'm testing out a fender bassman compact (early 80's soild state) and low and behold it pops the brand new arc fault during power down. What's even more strange is it only does it about 4 out of 10 times and happens on power up and/or power down and never while it is running. It doesn't do it through my isolation transformer/variac but that was kind of expected. Any thoughts? Also, secretly wondering how many amps I've worked on that had similar issues in my last shop while I was carrying on with your standard breaker setup. It does have a 3 wire cord and what appears to be a death cap. Guess I could start by removing the death cap and see how it responds. Good schematics for this amp seem to be scarce.

[g1]

The only schematic I found is pretty poor as you say. One thing I do notice is the power switch switches both sides of the AC, which is less common in North America.

[Enzo]

Your 120vAC mains sweep from zero to +170v through zero to -170v, and back to zero. Over and over. When you flip the power switch, the voltage at that instant could be zero, could be +170v, could be -170v, and could be anywhere in between. If I had to guess, I'd suspect your GFI is detecting the arc/spark when the switch flips at one of those 170v extremes. But not when ti switches near zero.

[TomCarlos]

What would happen if you put an Isolation Transformer between the amp and GFI? Or what about an APC backup battery unit?

I found an interesting read on the "Gear Page" - Amp trips breaker when you turn it on

[The Dude]

......Guess I could start by removing the death cap and see how it responds......

That is a good strategy.
It should be removed anyway.

[tdlunsfo]

Wonder if the ol' cap or resistor across the power switch may help.

[Pixel]

Does it happen if you turn it off from another switch? Try plugging it into a plain old power bar.

[Greg Robinson]

The proper way to diagnose this issue is to use an insulation resistance meter (Megger).
A GFCI/GFI/RCD trips when it detects an imbalance between active/live/line/phase and neutral/return. Usually any imbalanced in a class I appliance (grounded chassis) is through the protective earth conductor, though it can be through other paths to earth (like pipes, damp floors or your body) under certain fault conditions.
Inrush current, as had been suggested by some other posters, will not cause a GFCI to trip, unless it is a RCBO (combined GFCI/CB with overcurrent protection). This type will often have some indication of which fault caused them to trip - ground fault or overcurrent, so make sure to look out for that. Also make sure that it is in fact a GFCI that is tripping and not a regular circuit breaker.

An insulation resistance meter measures resistance by applying a high DC voltage. To test an appliance, this voltage is applied between the active/neutral part of the primary circuit (fuses must be present and power switch must be on) and chassis/protective earth. The high voltage helps to detect any insulation breakdown that a low voltage test like mulitmeter continuity will not detect. Such faults may be primary-to-secondary insulation breakdown, primary-to-core insulation breakdown, or other insulation breakdown in the primary-side circuitry, like power switches, fuse holders, TVS devices like MOVs, primary side caps, or just the wiring itself.

Any resistance lower than a couple of megaohms between the primary circuit and protective earth indicates a problem, use a process of elimination to find the cause.

The "death cap" shown on the North American variants in the schematic shared by G1 is a legacy of American ungrounded outlets, and is not required when using a grounded outlet (in linear 50/60Hz circuits like this, quite different case for the primary-to-secondary class Y capacitor needed for common-mode EMC supression as seen in SMPS supplies), so feel free to remove it.
If you decide that cap should remain, it must absolutely be a class Y type safety capacitor. This capacitor standard was developed exactly to prevent this type of fault, and any primary-to-ground capacitor must now be of this type by law (there are also maximum allowed values to limit AC earth leakage - this varies wildly by class of appliance and jurisdiction, check local guidelines), but this design is old enough that it is not a given that this type was fitted at the factory.

[g1]

If you decide that cap should remain, it must absolutely be a class Y type safety capacitor.

It may be very unpopular for me to say so, but this is not true everywhere. Maybe for manufacture, but not for repair of existing units. Depends where you are I guess.
Perhaps you were just expressing your opinion and not speaking in legal terms.

[Greg Robinson]

It may be very unpopular for me to say so, but this is not true everywhere. Maybe for manufacture, but not for repair of existing units. Depends where you are I guess.
Perhaps you were just expressing your opinion and not speaking in legal terms.

In almost every part of the world it is in fact a legal requirement. But like I said, consult local guidelines.

[Jon Snell]

Remove the death cap, if USA and your problems are most likely over.

[tdlunsfo]

Just out of curiosity, I replaced the death cap with another I had lying around to see if it would stop tripping the ground fault breaker and it didnt. I then removed the cap (which I intended to do anyways) and so far no trips. Interesting. I guess I have two leaky capacitors or a new mains circuit sensitive to line to ground capacitors, which isn't a bad thing I suppose. I guess I need to stock up on X and Y rated caps.

[Jon Snell]

They are probably not leaky, just dangerous. They conduct a small current, as they should when the AC is changing polarity, that can cause the trip to go.
We have Residule Currend Detector trips in the UK and they work on the same basis.
Line supplies the current and neutral takes it back (as a simple explanation) through a balanced electromagnet. If it becomes unbalance, (because the return goes slightly to ground), a magnetic field is built up, triggering the release mechanism.
Old hat but effective.
This may interest you; https://www.youtube.com/watch?v=XTUUijDclWk

[g1]

Just out of curiosity, I replaced the death cap with another I had lying around to see if it would stop tripping the ground fault breaker and it didnt.

I'll wager that a class Y in that position will still trip the GFI.

[Richard]

Can we get clarification as to whether the breaker was AFCI of GFCI? I saw both terms used in this thread and the terms often get comingled. To the best of my knowledge GFCIs are required where water is likely, bathrooms, kitchens, and garages. AFCI is required everywhere else, at least in the USA. A shop could be a garage or not.

For those just tuning in, GFCI has been in wide use for over 40 years. It detects imbalance in current between the hot and cold wires and assumes you are using the hair dryer in the shower again and kindly kills the power, allegedly before you could feel a shock. I first became aware if AFCI about 10 years ago, it's probably been code for 15 years. AFCI detects intermittent current of an arc and assumes your house is about to catch fire.