Brushless DC motor testing techniques

[mr.electric]

I just went to an advanced Prius repair class and found that my e-bike experience was very helpful in trouble shooting the three phase brushless dc permanent mag motors in a prius. We looked at a Gen 1 prius with a motor that had two phases shorted. You can roll the car in neutral and feel the cogging of the motors and the lumpy drag of a shorted motor. The problem is finding a partial short. The recommended technique is to put a dual trace scope on two phase wires and roll the car. You can easily see one smaller wave form coming from the bad phase. Another technique is to check the impedance of the windings with a milliohm meter. The problem is scopes are $1000 and milliohm meters are $800. Any alternatives for checking phase coils that are shorted turn to turn but not open circuited?

 

[gwhy!]

there are some very good,very cheap hand held lcd scopes that are on the market, you dont have to pay $1000 for one for this sort of work.

 

[amberwolf]

Heck, at the frequencies thsi stuff runs at, my old all-tube scopes work fine. Even if they do weigh more than the whole bike. :lol: So you can find any old dual-trace scope to compare waveforms.

 

[mr.electric]

Good point, I should start looking outside of my normal tool suppliers. I found some scopes for about $400. http://www.amazon.com/Precision-2120B-20mhz-Probe-Trace/dp/B000F0GM8E/ref=sr_1_fkmr0_3?ie=UTF8&qid=1284579557&sr=8-3-fkmr0
Still this is a big expensive piece of equipment to add to my tool box.
I was just wondering if there is another test besides using a scope

 

[johnrobholmes]

To measure the milliohms of the phases on the cheap, get a charger than has a motor break in or hot wire mode. Some chargery models have that, plenty of other brands too. Set it to regulate only 1 amp of current through a motor phase, then put a volt meter set to millivolts in parallel with the phase leads. The charger will throttle the voltage to allow only 1 amp, and the voltmeter will show the millivolts easily. 1mv = 1mO, super cheap and accurate enough for horseshoes or hand grenades.

 

[fechter]

To find turn-to-turn shorts in a winding, you really need a LCR meter. If you search eBay for "LCR meter", you'll see a wide variety. Here's a cheap one: http://cgi.ebay.com/Portable-Handhe...938?pt=LH_DefaultDomain_0&hash=item20b101c60a
$130 buy it now price.

This also measures milliohms quite nicely.
I have a B-K version of this and it's very versitile, but much more expensive.

 

[mr.electric]

Is it possible to test the phase coils by making a circuit that passes an ac current through the different windings and compares the results.

 

[mr.electric]

I see that an lcr passes an ac current and compares " phase angle between voltage and current" I imagine in practice it gives a number to each coil which should match within a certain tolerance for all three phase coils. Kind of like an electric motor compression test.
A tester that compares the three phases and gives three LEDs when all three match would be useful but it seems like a very complex circuit judging by the cost of the meters.

 

[fechter]

Yes, they are fairly expensive, but the best way to test a coil. If you wanted to make a tester that was only going to be used on one type of motor, you could possibly make something cheap that did not have a wide measurement range.

If you have a winding with, say, 1 turn shorted out, a DC resistance measurement might not show much difference even if you could accurately measure that low of a resistance. You really need to put AC through it to find a shorted winding.

Using a LCR meter, you could simply compare the readings from all 3 phases and they should match. It doesn't take long to take a measurement.

About the only other way to test would be to see how much drag there is aginst spinning the rotor. With even one turn shorted, there would be significantly higher drag on the rotor. This would reqire spinning the rotor somehow.

 

[texaspyro]

If you have a winding with, say, 1 turn shorted out, a DC resistance measurement might not show much difference even if you could accurately measure that low of a resistance. You really need to put AC through it to find a shorted winding.

Using a LCR meter, you could simply compare the readings from all 3 phases and they should match. It doesn't take long to take a measurement.

Depending upon the motor or transformer, a single shorted turn can be difficult to see even with an LCR meter. Probably the most sensitive measurement to shorted turns is the coil "Q", which any decent LCR meter can measure.

If you have a signal generator and scope you can try driving the coil with a square wave and look at the ringing voltage waveform on the falling edge of the waveform. Each phase should look identical. Shorted turns tend to have fairly dramatic effects on the waveform.

 

[mr.electric]

Interestingly you can do the drag test by pushing any Toyota type hybrid while it is in neutral. This includes all toyota, nissan and ford hybrids. Pushing in neutral turns both MG1 and MG2 the two motor/ generators in the trans. It is best to disconnect the phase wires first before doing the push test to rule out a short in the inverter. We want to be able to test for shorts more accurately than the push test. There is a check engine light for magnetism (weak magnets) and short to case but shorted windings makes check engine lights that correspond to inverter appear. Which in turn results in good inverters being replaced.
"A test is worth a thousand opinions" - R. Fechter