LMS Sieg SC4 Popping My Ground Fault Interrupter

[CFLBob]

That's an interesting story! I'm pretty sure the Micromark products are made by SIEG, too. Probably similar motors and controllers - scaled to save a few pennies for the smaller ones (or made for the smaller and used on the bigger ones to save some money).

My SC4 was five years old in December. Due to some problems with fixing other things in the house, the last time I used it extensively was in the summer. It's possible it's starting to go bad, but when plugged into a plain grounded outlet - like manual says - it seems to work fine.

 

[jjr2001]

Yes, I believe SIEG makes them under a lot of brand names. Great that it works on non GFI.
Cheers, JR

 

[kf2qd]

You might try putting a couple of ferrite chokes on on the power cable. You are seeing induced noise on the ground wire.

 

[awake]

I'm discussing this on another forum and somebody pointed out this quote from the LMS3540 product page on Little Machine Shop's web page It's on the tab marked "Chris' Tips"

We have no expectation that any HiTorque machine will work on a circuit with a Ground-Fault Circuit Interrupter (GFCI). The National Electric Code has a specific exception to the requirement that receptacles in a garage or work area have GFCI: "Exception No. 2: A single receptacle or a duplex receptacle for two appliances that, in normal use, is not easily moved from one place to another and that is cord-and-plug connected".

I think you just found your answer: apparently, Sieg has designed the PWM controller in such a way that it is electrically "noisy."

Which does and doesn't surprise me. When I blew the controller on my little 7x14 Grizzly, I did a good bit of tracking down and puzzling out of the controller circuit. It was pretty ... basic, shall we say. No isolation whatsoever, even/especially on the logic circuitry. The mains are rectified to DC - that part is not surprising. But then a resistor divider was used to generate the logic circuit voltage (!), and it did so on the high side of the DC rather than the low side. Thus, compared to the mains ground line, the logic ground was > 100v DC! Presumably their goal was to switch the N-FET on the high side rather than the low side. Ironically, the circuit used an opto-isolator to drive the FET, but that didn't actually do much isolating, since the logic circuit was tied directly into the high voltage DC.

I am not an expert with electronics, just an experienced hobbiest, but it all looked rather suspect to me, and I can imagine that it would lead to a very noisy circuit on whatever line it is plugged into. FWIW, I made my own controller, isolating the logic circuitry from the mains with a small transformer. For many years now it has performed flawlessly on the GFCI circuits in my garage. Of course, that might just be beginner's luck!

I could hope that the newer Sieg machines and their clones would be using a better controller than the original on my 7x14 ... but maybe not?

 

[Wizard69]

I just logged in and saw this thread.

first off the behavior of GFI outlets do vary a bit based on the manufacture. You could very well have a sensitive GFI. I tend to doubt that at the moment based on other info in this thread.

second if the GFI is protecting downstream outlets and you have stuff plugged into them then you really should unplug everything. The problem here is that leakage currents add up, so every thing plugged in can potentially be leaking.

some time ago I had real GFI issues and traced it down to a doggy extension cord. Note that from the outside the cord looked practically brand new. However it leaked in some manner and would trip GFIs. At least I believe it was leakage the cord did not show signs of being shorted.

The information that Little Machine Shop supplied is likely accurate. It isn’t uncommon for noise mitigation techniques to leak current to ground. This depends on the design of the electrical controls. This is not a problem on equipment that is properly grounded. However you need an outlet that isn’t GFI protected.

 

[Ken I]

Just to throw an alternate hypothesis in here.
GFI's will trip on a neutral to earth leaks as well - neutral and earth are at approximately the same potential (theoretically - they never are because of volt drops) - so if you have a neutral to earth fault at your machine and you pull 1A on live then 0.5A goes back through the earth and 0.5A through the neutral - so the GFI "sees" an imbalance between live and neutral and trips.
Neutral to earth faults can normally be deduced because it doesn't matter what you switch on (on that circuit) it causes a trip.
But the fault may not be a dead short only a resistive fault and that will only trip when a certain load creates sufficient volt drop between earth and neutral (at the fault) to bring about the necessary current imbalance between live and neutral to cause the GFI to trip. And a highly resistive connection (like a wet fault) will need a much larger load to create the imbalance - so you might be working quite happily and someone elsewhere in the house turns on something and the trip occurs.

Bear in mind the fault can be anywhere in your system related to that GFI it doesn't have to be anywhere near the device that appears to "cause" the trip when it is switched on.

Neutral to earth faults can be very difficult to trace - my tip would be to disconnect all loads on the circuit - disconnect the neutral from the board and start hunting with a meggaohm meter.
With Neutral off the supply, the neutral line should be as clear of earth as the live line.

Once you have given your system the all clear - then start hunting for the same problems in all your appliances - and as Wizard69 pointed out an extension lead is an "appliance".

Also your neutral to earth voltage (under nominal current load) should never be more than a few volts - 30V is way too much and could be the result of a hot joint or undersized wiring for the load.
This might also be indicative an induced voltage in an earth loop.

A hot joint / wet wiring etc can all lead to GFI trips.

RF interference (from your invertor) can also induce currents in "Earth Loops" that might cause trips - try to avoid Earth Loops - all appliances must be earthed to a single earthing point in the board.

An example of an earth loop would be a lathe at one end of a steel bench and a mill at the other end - each with its own earth but also grounded to the common bench - thus the two earth leads and the bench form a loop - or an antenna for the RF to induce into.

If I get high spurious earth neural voltages I would normally slap on an oscilloscope to see what I'm dealing with - sometimes induced interference from inverters etc. causes harmonics which can also create havoc with GFI units. Cycling the various inverters up/down / on/off normally identifies the culprit.

Regards, Ken

 

[CFLBob]

Thank you @Ken I and @Wizard69. The extension cord story is ... frightening, to be honest.

The sum of the last few days is that I moved the lathe from the west outlets, which have a GFCI, over to the east group which doesn't. The lathe is completely usable again. The manual doesn't call out having a GFCI, just a grounded outlet, which it has.

The lathe is just the latest episode of problems I've been having with the west branch throwing a ground fault and popping the interrupter. The random GFCI popping goes back not quite as long as the life of the addition but almost as long. I can't tell you how many times I've changed something and we've gone weeks without the thing popping and suddenly it does it again. The reason I had replaced the GFCI when I stumbled across this problem was to try to fix random popping.

Yesterday my wife was out there getting something and the GFCI popped while she was in the room. There was nothing turned on or running at the time. That's how it usually is when the GFCI pops. Ordinarily, I go out first thing in morning and find it. Sometimes it's not overnight, it just pops when no one is out there. With the exception of last week, when I found I could make one pop deliberately, I think I've only seen it go twice since May of 2014 (when the addition was finished). It has shut off my CNC equipment during a job exactly once. Conveniently, I had just zeroed everything, so it didn't mess up any work.

I've never bothered to write down all the things we've tried, but yesterday I had unplugged all but two things: a switched outlet with no surge protection in it (my CNC computer and big control box are switched by that) and a surge protected strip that has several low-current loads in it (my three Sherline DC motors, two CNC controller boxes, some lights, and an exercise bike). All of that was off yesterday when it popped.

A few years ago, when I found my house ground rod had corroded open and I put in a new one, I thought we were over the troubles. Since then, I've done other tricks that seemed to make the problems go away; replacing the surge protected AC strips with unprotected ones, swapping the east side GFCI with the west side's one, other things I forget.

So far the only thing I haven't tried is to unplug everything in the room when I'm not using it. It would be a bit inconvenient to have to plug everything in every time I use it, but it might restore some of my vanishing sanity. If it works.

I suppose this is probably far enough topic off now to not belong here, since it's about my five year old problem and not the lathe, but suggestions are appreciated.


Bob

 

[ignator]

Bob, your problems all sound like they are from the common mode EMI/EMC filters that are used to protect the electronics in your lathe, as well reduce RF emissions per FCC requirements. I could not find a schematic of the electronics for your lathe on line, but with a brushless motor, I suspect the input PI filter, where the capacitors are connected to ground, not neutral. Hence the requirement for a grounded outlet.
The other possible issue, your GFCI may be wired incorrectly, as I just found one, where the original electrician swapped wires from the input power to the protected outlets beyond, the neutrals were switched, and it randomly would trip. Easy to check this by looking at where the wires are in the bottom of the device box.
Your computers have this same input PI filter arrangement. It all depends on the capacitor size on how much current is dumped at 60hz to ground, and if this is enough mismatch for the current transformer in the GFCI breaker to trip on.

 

[CFLBob]

Thanks, ignator.

I understand your point, but since it's all 120V single phase, the filter is probably on the side that's switched and not in-circuit to the GFCI. When I replaced the interrupter outlet last week (three times), I verified the line and load connections to the outlet were proper by only wiring the line side and then turning the breaker back on to verify that voltage was on the outlet as it should be and the downstream outlets were off.

I've resolved the question of the lathe by simply plugging it into the side of the room that doesn't have a GFCI. We took that side's GFCI out years ago because we put in a freezer and had read about refrigerators and freezers causing them to pop.

I think I'm going to adopt the habit of unplugging everything when I'm not using it. That way if I go a long enough time, say until the summer thunderstorm season starts in May, without it popping, I can reason that it's either in the wiring or it's the power company's feed to us.

 

[SailplaneDriver]

ignator has it right. It could be the filters in the drive. I'm a retired electrical engineer that specialized in power system anaylsis and protection. Here are some other things to keep in mind and things you may want to check:

- GFCIs in the US are designed to trip between 4 and 5 miliampers of current. It doesn't take much to cause one to trip.
- The GFCI measures what goes out the hot wire and compares it to what comes back through the neutral (white wire in US). The GFCI will trip f the difference is more than the 4 to 5 miliampere setting.
- A connection of the neutral to earth anywhere after the GFCI device can cause some neutral current to bypass the GFCI causing nuisance tripping. Isolate the neutral at the GFCI device and test for continuity to ground (will all devices unplugged from the circuit) to see it that is causing a problem.
- Conservatively, 100 feet of wiring at 120 volts 60 Hertz has enough capacitive leakage current to cause 5 miliampers of current through earth return. Much more than 100 feet and you are asking for trouble. This is a big issue for pools and the protection of underwater lights. It can also be an issue for extension cords.
- Most residences only have one GFCI that covers the garage, bathrooms and outdoor receptacles. The total wiring can be 100 feet or more and make the circuit prone to nuisance tripping.
- Moisture in an outdoor receptacle on the GFCI circuit can cause leakage that will make the GFCI circuit prone to nuisance tripping.
- Receptacle GFCIs have less wiring after the protection than circuit breaker type GFCIs. The less wiring makes the receptacle type GFCI less likely to nuisance trip. The receptacle type is also much less expensive.
- Noise filters connect to earth and will add to the normal wiring capacitive leakage current.
- Many plug strips include filters and will add leakage current. Too many devices with filters can make a GFCI prone to nuisance trip.
- The National Electrical Code (for the US) requires all circuits in the garage to be GFCI protected under 2020 Article 210.8(A)(2). There used to be an exception for dedicated appliances such as freezers and garage door openers but that appears to have been removed.
- Using an adapter to isolate the ground of the lathe, or any other load, is extremely hazardous. If the device shorts to ground you will become the return path to earth. Only do this for testing and be very careful. Test the frame of the device to earth to make sure it is not energized before touching.
- GFCIs are electronic devices and will eventually fail. Test periodically to make sure they will trip. Use the test button on the unit. Better yet, get a receptacle tester with a ground fault test button on it and use that. This will not test if the device is too sensitive. It takes more specialized equipment to do so.
- Some GFCIs are sensitive to harmonics - the hash on the sine wave caused by electronics.
- Some GFCIs are sensitive to the starting current of motors which may cause them to nuisance trip.

I don't know of a simple way to test the filters for leakage. If you have a multimeter that can read AC amperes, you can make a cheater cord that will break out the earth return. Plug the lathe into the cheater and the other end into the mains receptacle. Run the cheater earth return through the meter and read how much current is returning to earth. If it is above 5 miliamperes, the problem is in the lathe somewhere - filters or some sort of short. If the current is well less than 5 miliamperes, the problem is elsewhere. If the current is above about 3 miliamperes, the combination of the lathe and leakage current from the wiring and other devices on the circuit is likely the issue.

GFCIs are simple devices and work well when properly installed and without excessive electronic devices plugged in. Check the above items to see if any of them could be the issue.

 

[CFLBob]

As my profile says, I'm also a retired EE, but my specialty is radio frequency design: receivers, transmitters, PLL frequency synthesizers, feedback/control loops. Anything in the DC to microwave spectrum. Consequently, I understand how they're supposed to operate, what I don't have any feel for is how tolerant they are of things that designers didn't design specifically for. In my last 20 years, I worked in commercial avionics. Everything electronic for the big air transport airliners and the smaller business jets. In aviation, everything is tested to major industry specifications so you know that nothing in the expected environment will cause the equipment to misbehave. I know nothing about how GFCIs are certified but I bet they don't go to that level.

At this point, I'm not thinking about the lathe, but about the fact that I've had random GFCI pops for over 5 years now and I'm about sick of it. We have other GFCI outlets in the original house and a GFCI breaker in the house breaker panel. None of those pop at all. Ever. (Except when we test them). The one spot I'm fighting in the shop has never gone more than a couple of months without the GFCI popping. The brand new 20A GFCI randomly popped yesterday.

Virtually all of the hundreds of times that GFCI popped, everything in the shop was turned off. As I said above, what made the lathe unusual was that I could make it fail just by increasing RPMs.

About the "100 feet of wiring at 120 volts 60 Hertz has enough capacitive leakage current to cause 5 miliampers of current through earth return." The shop is divided in half: east and west breakers. The breaker box is connected to a GFCI at the first outlet, maybe 20 feet away, and that outlet protects the rest of the circuit. From the breaker box to the last outlet is about 60 feet if you include going up into the attic and back down to the outlets. This isn't just a straight series connection of 60 feet, it's some up, some down, some series and some parallel. It's all "Romex" type, plastic-jacketed, three 14 AWG conductor wiring. The west wall (the problem wall) also has an outdoor outlet, a branch of about 12' of extra wire.

Both walls had the same arrangement, but I took out the one on the east side when we put the big freezer on it.

Technically, I don't believe this is a garage. Does that affect how the NEC considers it? There's no space for cars, no doors big enough to allow a car in if I wanted one. The contractor said we shouldn't call it a shop because that might bring county ordinances for workplaces into consideration. It's a large, open space, 770 sq. ft. It could have easily been more living space for the house with internal walls and (probably) plumbing.

My inclination is to remove the outdoor outlet. We were initially suspicious of it because of moisture, but there's no correlation of the GFCI popping after rain. Went down that road years ago.

 

[awake]

Oops, missed that detail in your profile (EE). I'm guessing you know way more about this stuff than some of us who have offered advice! Which makes your 5-year saga even more puzzling ...

Wish I had something useful to offer, butt I will certainly be interested to hear the details when you do figure it out!

 

[CFLBob]

Well, for all the good it has done for me, the background doesn't matter.

Troubleshooting intermittent failures is hard. With radios we try to get them to fail constantly instead of intermittently. If they fail cold, we spray them with refrigerants. If they fail hot, we use heat guns (like paint peelers) and if they fail under vibration, I've been known to whack them with a rubber mallet.

This system fails randomly at unpredictable intervals. Like I say, I've only been in the room when it popped a few times. There's probably a dozen things we did that were followed by not failing for a couple of months. Now I think all of it was totally random.

 

[SailplaneDriver]

Removing the outdoor receptacle is a good start. Do the continuity test to ground on the neutral wire. I've seen lots of problems in industry because of improper grounds on neutrals below the ground fault protective device. The same thing will happen to your GFCI if the neutral has an accidental earth downstream. Also check the earth return current on the lathe. That will pick up any problem in the lathe itself.

One of the guys I used to work with told a story of how he went to investigate some electrical switchgear that had burned down. He met the electrician on site. The electrician explained that this one circuit breaker kept nuisance tripping. The electrician had taken the cover off to obseve the operation of the breaker. Each time it tripped he saw a little lever jump up - so he wired the lever down. Turns out that there was an intermittent short circuit below the breaker. The electrician had disabled the trip unit by wiring the little lever down. The breaker overheated from the short and failed resulting in all the switchgear burning down. Sometimes when somethning trips it's trying to tell you something.

 

[SailplaneDriver]

As to wheter your shop is a garage or not, if it is at or below grade and not intended as habital space, it has to have GFCI. See the snip of the 2020 NEC below. Depending upon when your shop was built, you may be grandfathered to an earlier edition of the Code and this requirement may not apply.

 

[CFLBob]

As to wheter your shop is a garage or not, if it is at or below grade and not intended as habital space, it has to have GFCI. See the snip of the 2020 NEC below. Depending upon when your shop was built, you may be grandfathered to an earlier edition of the Code and this requirement may not apply.

View attachment 113152

Well, it's certainly an accessory building. I'm not aware if the Powers That Be here require everything to be retrofitted when the NEC is revised. I will have to talk to the building inspector's office.

 

[SailplaneDriver]

Well, it's certainly an accessory building. I'm not aware if the Powers That Be here require everything to be retrofitted when the NEC is revised. I will have to talk to the building inspector's office.

You are generally "grandfathered" and not normally required to update to current code unless you make changes. I would not bring it up to your building department since it could open a can of worms.

What year was your shop built and what city/state are you in? I can look up the code requirements at the time of construction.

 

[CFLBob]

I'm pretty certain that the building doesn't have to meet the current code unless it has been modified and needs a new inspection.

The shop was built in 2014 and I'm in Melbourne, Florida.

One of the guys I used to work with told a story of how he went to investigate some electrical switchgear that had burned down. He met the electrician on site. The electrician explained that this one circuit breaker kept nuisance tripping. The electrician had taken the cover off to obseve the operation of the breaker. Each time it tripped he saw a little lever jump up - so he wired the lever down.

That's mind-numbingly stupid for an electrician. It would be pretty stupid for Joey Bagadonuts, random homeowner, but for someone who should understand why breakers are there...

Sometimes when something trips it's trying to tell you something.

That's always my assumption. I've never replaced a fuse with a penny - or copper bar.

 

[Ken I]

Sorry Bob - I also didn't see your bio EE - my apologies for teaching you to suck eggs (WRT to earth loops and RF interference).

I once had a problem which was ultimately traced to the nearby electrified railway having lost one of its local earth spikes due to corrosion. The system was thus grounding HT literally through the earth. Since it depended on how dry the ground was (seasonal) and how heavy the train was and whether it was accelerating or not as it passed by - it was an infrequent cause and not easily identifiable. Eventually we realized it mostly happened in the dry season and finally when a train was stopped by a red light and had to pull away again. The rail company fixed it as soon as we notified them that we thought they had a problem and GFI faults stopped.

As Sherlock Holmes says - when you have eliminated all possible causes, then only the impossible remain.

Good Hunting - Ken

 

[CFLBob]

Again, no problem, Ken. For all the good it has done for me, the engineering background is hardly worth mentioning.

That ground issue you talk about is one of things I'm trying to figure out how to assess. There are tracks about 6/10 of a mile from here, probably too far.

Among the things that I've thought would have fixed it for good was when the power company replaced a transformer we share with three other families. We have underground utilities here so it's not branches touching a wire and interrupting it. We started to go through a period when the lights would dim or pulse in brightness and that was highly correlated with the GFCI popping. We called it in and after some troubleshooting, the power company came out and did the replacement. It was about two months before it popped again, and I really thought it was going to stay working until then.

Maybe I should try some of the troubleshooting things I'm used to on the outlets. Hit them with a heat gun and see if I can cause one to fail. It has never correlated with cold, so freeze mist seems like the wrong direction.