Chop saw tripping other gfci’s

[hillbilly1]

Had a strange one the other day, LED highbays off a 100 amp temporary. Four circuits with four GFI’s. Two on one leg, two on the other. Chop saw on a 100’ 12 gauge cord on one circuit, Five 240 watt LED’s on the another 100’ 12 gauge cord. As soon as the chop saw is started, the gfi for the LED’s trips. Moved the LED’s to another circuit, still does the same thing, moved to the third circuit same thing. Trips about 90% of the time regardless of which leg or circuit it’s on. No voltage drop detected at receptacle using a Fluke 89. Fixtures are rated for 100-277 volts, running on 120 volts. Maybe a spike caused by the saw? Haven’t had that sort of problem before.

 

[hillbilly1]

Sorry, didn’t notice I was still in campfire chat. If one of the mods can move it to troubleshooting! Thanks!

 

[LarryFine]

Hmmm. Voltage drop on the feeder neutral?

Do unloaded GFCIs trip, too?

 

[hillbilly1]

No, just the one with the lights. The chop saw gfi does not trip. 120/240 open delta service.

 

[ptonsparky]

I haven't used that meter but I have the 87. Put the meter on the lighting circuits. Look for Min/Max when in the Peak mode. Both Line and load side of the GFCI.

I usually find the issue is when saw is stopping.

 

[hillbilly1]

Voltage sitting steady on Min/Max on saw startup, that’s what’s so interesting. 200’ #2 Al quadraplex feeding the temp panel. The LED’s and chop saw is the only loads. The panel it feeds from has only three 100 watt LED highbays, and is only 30’ from the service.

 

[ptonsparky]

Voltage sitting steady on Min/Max on saw startup, that’s what’s so interesting. 200’ #2 Al quadraplex feeding the temp panel. The LED’s and chop saw is the only loads. The panel it feeds from has only three 100 watt LED highbays, and is only 30’ from the service.

No VD?

 

[synchro]

Perhaps the following could be happening:
The current drawn by the chop saw could be drawing a current with a substantial broadband noise component, particularly if it has a universal motor with brushes. However, if the current it draws is just between line and neutral, and therefore without a significant common-mode component, then the GFCI feeding the chop saw should'nt trip. However, the noisy current could create a noise voltage on the panel bus feeding the GFCI breakers.
Now the LED high-bays may have capacitors between L-G and N-G to filter out the interference originating from their switching supplies, which typically would be operating at 20kHz or higher. But at 60Hz such capacitors would have a high impedance and so draw very little current to ground, therefore not tripping the GFCI. However, at the higher frequencies of the chop saw induced noise voltage on the bus, the impedance of the capacitors to ground in the LED lights would be lower and therefore draw more current, possibly then tripping the GFCI breaker that feeds them.

It would be interesting to check whether the GFCI breaker for the lights will still trip if the lights are turned off. If they have a N-G capacitor, it might conduct a noise current that trips the GFCI even when a switch on its line conductor is turned off.

 

[ptonsparky]

Perhaps the following could be happening:
The current drawn by the chop saw could be drawing a current with a substantial broadband noise component, particularly if it has a universal motor with brushes. However, if the current it draws is just between line and neutral, and therefore without a significant common-mode component, then the GFCI feeding the chop saw should'nt trip. However, the noisy current could create a noise voltage on the panel bus feeding the GFCI breakers.
Now the LED high-bays may have capacitors between L-G and N-G to filter out the interference originating from their switching supplies, which typically would be operating at 20kHz or higher. But at 60Hz such capacitors would have a high impedance and so draw very little current to ground, therefore not tripping the GFCI. However, at the higher frequencies of the chop saw induced noise voltage on the bus, the impedance of the capacitors to ground in the LED lights would be lower and therefore draw more current, possibly then tripping the GFCI breaker that feeds them.

It would be interesting to check whether the GFCI breaker for the lights will still trip if the lights are turned off. If they have a N-G capacitor, it might conduct a noise current that trips the GFCI even when a switch on its line conductor is turned off.

Yes, what he said. Should be easy for most of us to track to that down. No problem.

 

[kwired]

Is chop saw plugged directly into a receptacle that is pretty close to the panel?

If so does plugging it into same receptacle but using a 50 foot extension cord before plugging into the saw help lessen or even eliminate the problem?

That trick does often work with standard thermal magnetic breakers with a low magnetic trip setting, particularly Square D QO and HOM single pole breakers. But that is always when the breaker supplying the saw is tripping on in rush current. Still may be similar with some kind of interference or inductive kickback if tripping breakers for other circuits and the extension cord provides reasonable choke effect to lessen the issues.

 

[suemarkp]

You say the saw and the LEDs are both on 100' extension cords. Are those cords run parallel to each other? An inductive kick in the saw cord could be coupling into the LED cord and cause the GFCI to trip.

 

[kwired]

Is chop saw plugged directly into a receptacle that is pretty close to the panel?

If so does plugging it into same receptacle but using a 50 foot extension cord before plugging into the saw help lessen or even eliminate the problem?

That trick does often work with standard thermal magnetic breakers with a low magnetic trip setting, particularly Square D QO and HOM single pole breakers. But that is always when the breaker supplying the saw is tripping on in rush current. Still may be similar with some kind of interference or inductive kickback if tripping breakers for other circuits and the extension cord provides reasonable choke effect to lessen the issues.

Guess I can't read, you said this is on a 100' cord in the OP

 

[drcampbell]

Yes, what he said. Should be easy for most of us to track to that down. No problem.

If broadband noise and L-G capacitance is the problem, a powerline filter or an isolation transformer might be a solution.

I'm presuming that you just want it to work, and care little or naught about a complete explanation from forensic engineering.

An example of a powerline filter:
(although this particular device probably doesn't have enough attenuation in midrange frequencies (61-10,000 Hz) to be effective)

https://www.farnell.com/datasheets/2707707.pdf

 

[hillbilly1]

No VD?

Less than two volts at the receptacle, probably a little more at the equipment, but the 100’ cords are both #12.

 

[hillbilly1]

You say the saw and the LEDs are both on 100' extension cords. Are those cords run parallel to each other? An inductive kick in the saw cord could be coupling into the LED cord and cause the GFCI to trip.

Cords are run in two different directions, one lays on the floor, other is tywrapped up to the Z joists to the lights.

 

[synchro]

Just as a test you might disconnect the EGC that feeds the lights to check whether it still trips. If it stops tripping then it's very likely that filter capacitors from L-G and/or N-G inside of the lights are conducting noise to ground, and that's whats causing the GFCI breaker to trip. If it still trips then some other mechanism must be involved.

 

[cadpoint]

In addition to the excellent answers already presented.
With something like this one might be able to see something with a combination meter/amp clamp.
There's a plug slot on the bottom,



It's always a classic problem when a motor is stretched out on a 100'.
Just read'n capcatinious on the 100' itself might be the tell.
I've seen and experienced it too many times.

 

[hillbilly1]

In addition to the excellent answers already presented.
With something like this one might be able to see something with a combination meter/amp clamp.
There's a plug slot on the bottom,

View attachment 2563354

It's always a classic problem when a motor is stretched out on a 100'.
Just read'n capcatinious on the 100' itself might be the tell.
I've seen and experienced it too many times.

I had one of those about thirty years ago, Tandy brand. It had a cord that eventually failed.

 

[hillbilly1]

What’s funny is the chop saw never trips, it’s always another gfi. I’m putting up the other five tomorrow on another circuit. I will see what happens then. Fixtures are permanently mounted, just temporarily connecting them until I get the new 277/480 service hot.

 

[Amps]

What class are the GFI's? A or C?