Phase-to-Phase Resistance Testing of MV Motor

[LargeACMotorDude]

Hi everyone.

I’ve recently come across an interesting test result on a 1000HP 2P motor we rewound and are preparing to ship.

When taking phase-to-phase resistance measurements through the winding process (done before and after a global VPI) the resistance imbalance was less than 1% across all three phases. However, after reassembly and no-load testing, the resistance imbalance has increased to around 2% and is erratic when tested multiple times.

Assuming that the test equipment is accurate, my initial thoughts are that there exists some residual magnetism in the rotor or stator laminations that is affecting the resistance balance.

What are your thoughts on this? I appreciate any input.

 

[waross]

Are you measuring the winding resistance or the winding to ground resistance?
Are you measuring resistance or impedance?
How are you connecting the test leads to the motor leads?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[jpts]

Residual magnetic field is a DC field, and shouldn't affect on results. I would check the contact from measurement device to motor terminals.

Was the measurement done immediately after no-load testing -->Could small differences in winding temperatures between phases explain this? Although it will not explain "erratic" nature of the results (again, I would check that contact resistance from measurement device to motor terminals is not distorting the results)

 

[iop95]

What device used for measurements?

 

[waross]

Are you testing winding resistance? (Best results with a four point measurement.)
Are you testing winding to ground?
Are you testing winding to winding?

Erratic how?
Are subsequent readings decreasing?
Are subsequent readings increasing?
Are subsequent readings random?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[LargeACMotorDude]

Waross,

We are testing the winding resistance between each lead (3 lead motor) using a Baker AWA. The AWA has three alligator clips for connecting to the motor ring terminals which were crimped to the lead wire and soldered.

The subsequent tests are erratic in that they are random, but are consistently balanced between 1.7% and 2.3%.

We noted a small amount of voltage generated in the windings when rotating the shaft (about 0.1 VAC). This further leads me to think that there is some magnetism left in the system that might be leading to the errant readings.

It should be said that when checking the resistance with a DLRO all readings are exactly the same readings on all three leads. It only seems to be an issue when trying to use our AWA. We are confident that the motor is electrically good ship but we are curious about where these readings are coming from.

Thanks for the replies.

 

[electricpete]

You might be thinking of the fact that DC measurements can cause residual magnetism which affects ac measurements (like meggering before magnetization test in a transformer can cause problems). However residual magnetism doesn't affect dc tests like winding resistance.

I guess there can be residual dc on the winding capacitance / absorbtance. After dc testing, a common thumbrule is that winding should be grounded 4x the length of voltage application to assure subsequent dc testing isn't affected by any residual dc from prior dc test (and possibly for personnel safety). I guess in theory something similar might (?) might happen after ac testing/running to a small extent, but I've never heard of it. At any rate after a first round of questionable dc tests, it certainly couldn't hurt to make sure winding is grounded for awhile before retestimg

Were the early measurements taken before lugging the cables and the late measurements taken after? (maybe one phase lug crimp is bad).

 

[LargeACMotorDude]

Pete,

That’s some good info, thanks for sharing.

The earlier measurements were taken before the terminals were installed. However, using two of our other measuring instruments we are not seeing the resistance imbalance we are getting on our AWA. We verified that the AWA is still measuring accurately using three other stators of known resistance. Perhaps it’s something in the method the AWA uses to take its readings?

We tested the motor resistance after being grounded for a full 14 hours and it didn’t seem to have any affect on the measurements. That was something we initially considered though.

Thank you.

 

[electricpete]

Sorry I didn't see the post just before my last post.

> It should be said that when checking the resistance with a DLRO all readings are exactly the same readings on all three leads.

If it's like our DLRO, I assume it's the exact same reading down to 3 decimal places (like 0.XYZ ohme) and taken with Kelvin probes. Like you, I'd trust it more than the AWA, but perhaps only because I'm not familiar with that AWA.

Does the AWA also use a Kelvin probe (for 4 wire resistance measurement to eliminate test connection resistance effects). Is the insulation between 2 jaws of each Kelvin clip still intact?

IF that AWA tester has 3 test leads that you hook up at once (PDMA does that, I'm not sure about Baker/Megger AWA), then it might be interesting to rotate the test leads and see if the oddball result tends to follow one test lead around.

 

[waross]

Not 100% sure but it looks as if the tester is using a 4 point measurement.
If the test clip is on stranded cable, you may have some inter strand resistance throwing off the results.
Both sides of the test clip must make good contact.
One side of each clip will inject a current through the winding under test. The other side of the clip will measure the voltage drop across the winding.
If both sides of the clip are not making very good contact with the lead, or if you are clipping onto stranded cable, then 1% or 2% may be measurement error.
Be sure that both sides of the clip are clean and that both contact points on the motor lead are clean.
With the very low winding resistance a slightly poor or dirty connection that would not make a difference on other higher resistance measurements may show a slight error.
Even removing a clip, rotating it 180 degrees and re-installing it may make a slight difference.
Try pinching the alligator clips tighter by hand and then moving them slightly to scrub through any surface contamination.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[LargeACMotorDude]

Yes the AWA is a 4 point measurement using 3 kelvin clips to take the readings. The alligator clips of the tester are clamped on the barrels of the ring terminals which have been sanded and cleaned.

We actually tested this motor with a PDMA as well which it sounds like you are familiar with. The difference between the AWA and the PDMA resistance testing is that the PDMA utilizes a ground lead to discharge the windings between each phase when measuring resistance. Using the PDMA, we did not see the higher disparity we are getting with the AWA. We took this as more evidence that the error is coming from the test method of the AWA.

Interesting that you bring up rotating the test leads, we actually tried this when we first ran into issues. When we rotated the test leads the error remained between test leads 2-3 regardless of what phase we have it on. I’m not sure what to make of that. If it was an issue somewhere in the test leads themselves I would think we would be getting errors on all of our motors instead of just this one.

We also tried taking the measurements one phase at a time (manually moving two of the clips around the phases instead of using all three of the clips) and again this this resulted in perfectly balanced measurements. Very strange.

 

[LargeACMotorDude]

An interesting anecdote that may, or may not be, relayed popped into my head after we encountered this issue.

A few years ago we were using our AWA to conduct an incoming test on an assembled MV motor. The motor was passing every test (500V RTG, DV HiPot, and surge comparison test) but was unable to get a steady resistance measurement.

After scratching our heads for a few minutes, we noticed that while the AWA was conducting the resistance test, it was actually causing the shaft of the motor to rotate. The rotating shaft was preventing us from getting a steady reading.

During disassembly and further testing, it was noted that the motor was manufactured using magnetic slot wedges to bolster the air gap flux densities between the stator and rotor.

We contacted Baker/Megger about this and their answer was that the magnetic wedges were creating enough of a field to cause the rotor to move during testing.

Probably unrelated but I thought you guys might be interested in hearing this.

 

[electricpete]

> When we rotated the test leads the error remained between test leads 2-3 regardless of what phase we have it on.

That seems like pretty powerful evidence that the problem is in the test box / test leads.

> I’m not sure what to make of that. If it was an issue somewhere in the test leads themselves I would think we would be getting errors on all of our motors instead of just this one.

Hmmm, so we're looking for something about the interaction between the test leads (or test box) and the motor (it only shows up on one motor). One truly whacky thought pops into my mind. Dissimilar metals can create dc voltage by thermoelectric effect (which could in theory interfere with a dc measurement). Maybe the lugs of this particular motor are different enough material to create a temperature dependent voltage. So then why only one lead pair?.... I think the Kelvin probes have larger conductor for current and smaller for voltage... if those leads were reversed the test lead and termination would run hotter. Maybe one or two of them is reversed and it only shows up when you connect to lugs with dissimilar metal lugs like this motor (postulated). I dunno, that's pretty far out there.

Some practical troubleshooting if you have an extra set of 3 test leads to use with your AWA, trying them out is probably the highest payoff (even if it isn't the exact whacky problem I postulated, many problems end up being test leads).

Otherwise, some other ideas with less likelihood to show anything. Maybe do a thermography scan during the measurement to see if any of the leads look different temperature. Finally (hail mary), maybe you could try connecting to bare conductor if enough insulation is removed at the barrel or enough extra conductor is accessible to grab onto at the tongue end of the barrel. That would change the material at the other end of the contact which could make a difference under the wacky theory there is a thermoelectric voltage which is material dependent.

No doubt baker / megger has a lot more relevant ideas than me if you have access to them.

> An interesting anecdoate... a few years ago [another motor] we noticed that while the AWA was conducting the resistance test, it was actually causing the shaft of the motor to rotate.

Whoa, who'd of thunk it! I guess the higher the current used formeasurement, the more likely to create noticeable magnetic forces. I think a DLRO uses 10A, maybe AWA uses similar.

 

[Compositepro]

"Assuming that the test equipment is accurate."

That is a horrible and lazy assumption. In fact, you should always start with the opposite assumption. Note that almost all the questions to you are related to your test method. You cannot associate any meaning to a measurement without thoroughly understanding the test method. You are saying that you are too lazy ("busy") to take the time to read and understand the operating manual and theory.

I don't mean to be mean. You really just probably worded your question poorly, by ruling out the correct answers, but you are getting the answers you need anyway.

 

[waross]

A problem with two windings?
Check the probe that is common to those two windings.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[LargeACMotorDude]

Composite,

This was just an instance of poor wording on my part as you said.

Our initial assumption when we noticed the odd readings was that there was an error in either our setup or our equipment. We eliminated the equipment being faulty by measuring three other stators of known resistance and verifying its accuracy. However I would never rule it out entirely as a possible cause.

We eliminated everything we considered to be an obvious cause of the issue but continued to get erratic results which only added to our confusion.

 

[LargeACMotorDude]

Pete and Waross,

You’ve certainly given me a lot to think about. I greatly appreciate an outside perspective when we come across something unusual like this. I find great value in these forums of experts with many years of experience.

I will continue to try out your suggestions and I’ll report back if I come to any conclusions.

Thanks again.

 

[edison123]

May be the three other stators resistances are within the measurement range of Baker AWA and this motor resistance is out of its range? If DLRO is showing no deviations, then I see no issues.

Muthu

http://www.edison.co.in

 

[electricpete]

If DLRO is showing no deviations, then I see no issues.

I agree there's no concern about the motor, and by this point in the discussion I'm pretty sure op agrees, but is hoping to understand his test equipment performance better.

Assuming that the test equipment is accurate, my initial thoughts are

FWIW I see nothing wrong with presenting your initial thoughts in that matter. The fact that you highlighted your assumption (rather than making an unstated silent assumption) brings appropriate attention to that assumption. Later clearly you started considering test equipment as you dug into the problem.

You’ve certainly given me a lot to think about. I greatly appreciate an outside perspective when we come across something unusual like this. I find great value in these forums of experts with many years of experience.

I will continue to try out your suggestions and I’ll report back if I come to any conclusions.

Thanks. I agree it's helpful to have people to bounce things off when something doesn't add up. I see it's your first thread and hope you'll stick around with us here and let us know what you find on this interesting problem.