Lathe electrically "hot" due to VFD? WTF?

Definitely check or establish continuity between the VFD and lathe frame to the ground wire. It should be there already, though.

Depending on how old the house is and who did the work, someone may have cheated somewhere along the line and connected ground to neutral. This is not uncommon in old two-wire systems that have been wired by someone who knows the inspectors will check with a tester without looking further (Lookie -- the green lights go on!). Depending on how the rest of the house is wired and running you can have some significant voltage on neutral, leading to mild shocks.

In large industrial installations its not uncommon to have different grounds in different sections of a building be tens of volts different

-- I had a problem with this in a former workplace, where two such sections came together in one part of the factory floor; plugging diagnostic computers into one section and the system to be diagnosed in the other lead to the magic smoke coming out. I can't imagine that happening in a normal residential setting, though.

If you can figure out a way to check the potential between the ground pin and the floor it may be educational, although I'm not sure what you could do with it.

Just a data point for you. I just got my large mill power supply running. I get 21 volts D.C. between neutral and ground. I attached a ground rod to this main panel. I've never seen this large a value. If I touched neutral while standing bare foot on steel, I'm sure it would tickle.

Anyway, I'd suspect neutral is touching your frame someplace.

Karl

I'll bet you don't. I'd hook a long wire to a water pipe and measure between it and the VFD. If you're getting zapped, I'll bet there's at least a 40 volt AC difference between the water pipe and the VFD chassis.

Follow the ground wire back to the service entrance measuring the voltage. See if it drops to zero at some point. Be sure your multimeter is set to AC volts.

Pretty much. Florescent lights with old style magnetic ballasts would do the same thing. The frame of the light can float up from ground enough to give you a good tingle if not grounded.

Try to find an open ground first.

Good plan. Let us know what you find.

Good so far, but a separate ground wire from the lug to the lathe would be a good idea.

If you've had a long dry spell the ground resistance can increase a lot - I sometimes have to water mine (as part of the 3-monthly check I do on my shed electrics) to get a safe resistance...

If you want / need to measure it, get a low-voltage transformer and attach one side of the low volts winding to the earth rod, other to a "roving" earth rod that you hammer into the earth and water (or if the local elec. co, provide one, use that) - do this in a number of spots at least 20-30 yards away and measure the AC voltage and the current into the rod, V/I will give you the resistance, check your local electrical code for the minimum value - to save the transformer for next time you may need a bulb in series with the low volts if the earth is any good - DAMHIK ;)

Dunno about VFDs (mine is on a *short, thick* cable to a pair of local earth spikes), but when I play with the ham radio (which has high frequencies buzzing about, although a few octaves above VFD frequencies) I find the house earth is too long for satisfactory earthing and I have RF on the equipment cases... for a narrow range of frequencies a 1/4 wavelength wire (insulated at the far end) works to absorb it, being a dead short at the right frequency - not easy to do for a *variable* frequency drive, tho! The VFD output is a chopped square wave, so has some pretty high-frequency components at fairly high voltages that can be a problem to shield effectively and can't be filtered easily (the magic smoke leaks out of the VFD if the HF impedance is too low) - one of the reasons that cable choice is important when installing 'em.

High-resistance in the ground circuit would be my first thought, too

Dave H.

Had a house in MD, garage on one side of a stream house on the other side.

"Ground" for the house was about 80 volts different from "Ground" at the garage.

Dave

Tim, I will reply to other fine points that you made when I gather more data, but yes, we had a very long dry spell with a very minor rain a couple of days ago.

I will check, very meticulously, whether I have a continuous ground, first.

Ultimately, though, the question is, is that the motor that is leaking, and if so, isn't that the primary issue?

Well, whatever's causing the problem is the primary issue.

Do you only get zapped with the lathe turning?

With the VFD on no matter what?

Any time at all?

The further you get down that list and still say 'yes' the less likely it is a motor problem, and the more likely it's a really odd ground problem.

I'd suspect bypass cap current into neutral also. Can you run the VFD with the motor disconnected to check?

jsw

Yes.

No

Only when the motor is running.

Your safety ground system, somewhere, has a problem. it COULD be the ground for the lathe is bad, or it could be the entire ground for the whole house is going out. this is quite common. Our ground went out somev time ago, but fortunately, the whole feeder went out in the same event, so we lucked out and didn't fry all the 120 V stuff in the house. Our neighbor was not so fortunate, and a bunch of electronics got fried.

So, basically, trust NOTHING, assume NOTHING.

These are most likely NOT the problem. Turn the VFD and lathe on, and then just keep working back to the panel. You need a real ground reference, but as I said above, you can't trust anything. You may have to drop a bar in the ground to get a real reference. But, you may be able to hook an extension cord to a known-good outlet somewhere else in the house, and then use the ground terminal on it with a meter to probe around to find the open ground connection.

In many wood homes, you can have a totally screwed up ground system, and have NO IDEA anything is wrong. But, if all your other machines are OK (I would NOT go around testing them with the bare feet and wet finger technique!) then it is just something on the lathe. If your safety ground is tied THROUGH the VFD, that is a definite possible culprit there. Could be a misunderstanding of what terminal is what, or just a bad connection. And, just because you have one continuous "wire" from here to there, don't be absolutely SURE there is solid copper all the way. As I said, trust NOTHING!

Jon

Hmm, I didn't read this carefully before. Do NOT make safety ground connections through anything aluminum and not expressly designed for wire termination. Aluminum instantly builds an oxide layer, and that can be quite a good insulator. Anodized aluminum can withstand several thousand volts, and some anodized aluminum cannot be distinguished from plain. But, anything bolted to aluminum sheet or whatever cannot be trusted to carry a safety ground over time. Make a copper wire connection for your safety ground, with terminal lugs approved for electrical connections.

I'm just guessing here, but this could be it. It also could be that your VFD case is isolated from the inner safety ground frame, so you grounded the VFD, but not the plate.

Jon

No. All motors "leak" as they have a lot of wire in close proximity to the frame. The VFD makes this much worse, as instead of 60 HZ on the windings, you have extremely fast sharp-edged PWM pulses at 400 V. The wire to frame effectively creates a capacitor, and the faster the dv/dt, the more current flows through the capacitor.

If you only got zapped when the motor was running, then the ground of the lathe has to be open. If you got zapped ANY time you touched it, then it sounds like the house ground might be going out.

Jon

Even metal water pipe can be a poor ground, and if you have much current flowing it will damage the pipe at the joints.

I didn't mean to imply that he should bond the VFD to the water pipe. Just to use it as a ground reference while he looks for an open in the house and machine grounds.

If the joints are corroded, it isn't a good reference. It can get you killed when you don't read enough voltage to bother you, when it is much higher. Also, with a VFD there may be a lot of harmonics on the neutral and running a wire to a water pipe won't show them.

I have looked at everything very carefully and found a place where ground was not connected properly. Thanks to all. This lathe runs very well now, everything pertaining to variable speed works extremely well, VFD is working, etc. It cost me more to fix than I hoped (VS stuff), but the result is very nice. The last thing remaining is to make stops for the taper attachment.

Glad you found it -- it's really unpleasant when the electrical stuff starts biting.

I agree. So now I have the best of all worlds, VFD for running from single phase, and hydraulic variable speed drive for preserving mechanical advantage.

Tim, I will reply to other fine points that you made when I gather more data, but yes, we had a very long dry spell with a very minor rain a couple of days ago.

I will check, very meticulously, whether I have a continuous ground, first.

Ultimately, though, the question is, is that the motor that is leaking, and if so, isn't that the primary issue?

Make sure that ground and neutral at the plug are the same. Years ago had a problem with a disk drive failing to spin up properly, and they had left off the jumper at the pole between ground and neutral. Was 12 V between.