RELIABLE and CHEAP STATOR TESTING.

[posplayr]

EDITED: 7/25/2015 Most all of these clamp on meters are only intended for 50-60 hz; I mis read the spec and the 400Hz was for voltage not current. It should still work as Toms results indicate.

I have mentioned this a few times here and Tom203 has actually done it, but I don't think there is a lot of awareness. While checking the charging system on your bikes has gotten much easier and better (see The Quick Test and other updates to the New stator pages) Stator winding testing (see Phase B of the stator tests) is still a bit of a hit of miss deal. A positive tests result is inconclusive, and only a negative tests result is conclusive.

http://www.thegsresources.com/statorpapers.php

For example if you use a 3V battery in a Ohm meter to excite the stator to see if it is shorted and you find a short, the 3V was enough to puncture the insulation and yes indeed you can assume the stator is bad. On the other hand, if you do the tests and it comes back positive, you know nothing!
That is because at 10V or 40V or at 100V the insulation might actually break down and confirm the stator is bad. So until keep upping the voltage and get a bad result, you don't know anything(this is what a MEGGER is for;there are cheap ones of these as well on Ebay), unless of course if the tests is performed at full load.

Any of these things will probably serve it's purpose even those it is being used in a way that is well beyond it's stated capabilities.

http://www.ebay.com/itm/VICHY-VC60B-Digital-Insulation-Resistance-Tester-Megger-MegOhmmeter-Meter-/331174091046?pt=LH_DefaultDomain_0&hash=item4d1b839926

http://www.harborfreight.com/digital-clamp-meter-96308.html

The phase B tests are also relatively high voltage tests but without much current. The stator AC should be about 80 V (at 5K RPM) and compare that to the typical lowest voltage setting on a MEGGER of about 250V.

As is true of almost any electrical circuit, testing at full load is best as it comes closest to actual operating conditions; it usually does not suffer the same kind of gotchas of other more indirect tests. The main issue in the past is actually finding a cheap way to measure AC current in the statro windings. I have a rather expensive scope with current clamp which is how I have done many of my measurements, but it turns out that a clamp on AC voltmeter also does a good job for this application. So while it has been know, few have probably done it, but there is really no good reason not to as you can probably do the tests using a $14 HF clamp on meter.

I was hoping someone would have bought one of the cheap HF clamp on volt meters to see if you can detect loaded current imbalance in the stator winding. I know Tom203 confirmed his SH-775 SERIES mode operation using one he had, just not sure if the model/quality exceeded the HF models to the point they would not work for this application.

http://www.harborfreight.com/clamp-o...ter-95683.html

If you have a digital voltmeter and just ride around, carry the clamp on meter with you. If you notice the voltage is dropping, then your the clamp on to confirm imbalance and you can be pretty well assured that the stator is going out.

Actually the frequency range is pretty narrow 50-60Hz. 1K RPM is about 125 Hz so you are well out of the range of spec performance.

Frequency range 50-60Hz

This is the general shape of the measurements you should get as a function of RPM. Note you will measure much more current if you have a SHUNT R/R instead of a SERIES R/R (that is why stators burn in the first place). Since you are measuring the current in a single phase, you will also not be measuring total current. That will also vary depending up whether you have Delta of Wye winding in the stator.

Most GS are Wye (see the manual) and so the RMS current in each leg is 2/3 of the total current (66.7%). So what ever RPM amperage you measure on a leg multiply that by 3/2 to get total current. If it is not balanced it is more complicated and yours is busted anyway so I won't bother trying to figure that out.

StatorCurrent.jpg

[steve murdoch]

I am a numbskull when it comes to anything electrical so it will take me while to digest this thread.
But i just wanted to say thanks to Posplayr and Tom203 and others for their continuing efforts in keeping our electrics sorted.

[posplayr]

I am a numbskull when it comes to anything electrical so it will take me while to digest this thread.
But i just wanted to say thanks to Posplayr and Tom203 and others for their continuing efforts in keeping our electrics sorted.

Thanks but sorry if I was not clear.

It is simple,

The current Phase B tests involve disconnecting the stator to measure the open loop voltage of the stator (tests says >65VAC at 5000 RPM). You are basically looking for balanced voltages across each leg. But this open circuit and there is no real load that might help to stress the insulation breaking it down so it fails. A good fresh stator insulation would pass the test.

Instead under the new test, you are operating under as much load as the RPM would dictate. (no light bulbs, no load dump resistors) Under the new test using the clamp amp meter , You leave the stator connected to the R/R and you put the clamp on around each of the three legs on AC amps settings. The chart basically shows how much current you should get on each leg as a function of RPM. Bottom line is they should all match. If one is low in current, the stator is bad.

In summary, if one of your stator legs is going out and not generating the power of the others, the AC clamp on meter will detect that. This avoids the all to often problem, of shutting down the bike to cool, disconnecting the R/R and having the stator show no signs of a problem because it has cooled. Some guys have been unfortunate enough to have just "on the edge" failures that come in and out having them chase their tails for weeks. (Psyguy had this happen once).

With the Ac clamp you can detect the failure while the bike is charging, where as there has never been a way to do that before short of having a relativity expensive current probe. At $14 from HF it is well within everyone's grasp to just buy one of those instead of the $3 ones.

DO NOT BUY THIS!!!!

http://www.harborfreight.com/7-function-digital-multimeter-90899.html

SPLURGE FOR THIS.

http://www.harborfreight.com/clamp-on-digital-multimeter-95683.html

[tom203]

The unit that I used had no brand visible on it .Compared to other units, it is unusual in its current ranges - 2, 20,and 200 amps. I checked a bunch of 60 cycle ac stuff with it, both very low current stuff ( less than .1 amp) and some 15 amp power tools and it gave reasonable results current wise. I'm skeptical of how accurate it can be measuring current as frequencies rise above the typical 60 cycle range, but i'm convinced that it showed a significant stator current difference when comparing a shunt unit to the series SH-775 .
I have had an outboard voltmeter installed from my first adventures in charging fun, so weird stuff happens when the stator starts failing. Voltage might become obviously low or erratic or it might just level off at 13.5 no matter what rpm you're at- my last failed stator hung on like this, still producing enough to keep bike running ok for another 30 hours or so. Removal of stator showed it bbq'ed in one section.

[posplayr]

Anybody not sure what a clamping amp meter is here is a video. The author actually recommends not buying a combo clamp and meter, but he is probably an engineer and wants engineering quality measurements. Buying the clamp separate from his meter allows him to use his existing (expensive) meter and just get the clamp. But you will also hear him say, current clamps by themselves are not cheap (he is using Fluke products). My Fluke add-io device I bought 20 years ago was about $450 to add to my $3500 scope.

With the harbor freight combo meter that also has a specification out to the equivalent of 3600 RPM you are going to get a good indication of AC current.

To be clear, if you want to measure AC current on say an AC hairdrayer, you have to separate the two conductors in the AC cord so the clamp encloses only one of the wires (either will work but only one).

For stator wires, as long as there is some separation of the individual wires to sneak the clamp in you are good to go . This is a safe way to measure as you don't have to disconnect anything and no exposed conductors are required. The Clamp meter basically measures the magnetic field around the wire to infer the amount of current flowing. To get a voltage measurement you have to actually probe the conductors as you are probably already aware.

I guess I should mention that you could use a VOM that allows the AC current to actually flow through the device. But now you have to worry about overloading your meter, blowing it's internal fuse and making sure you have good connections to and through your meter. The clamp on version, can NOT blow a fuse from the magnetic field and they are generally rated into the 100's of AC amps.

https://www.youtube.com/watch?v=WDTVE7lxJQ8

[posplayr]

The unit that I used had no brand visible on it .Compared to other units, it is unusual in its current ranges - 2, 20,and 200 amps. I checked a bunch of 60 cycle ac stuff with it, both very low current stuff ( less than .1 amp) and some 15 amp power tools and it gave reasonable results current wise. I'm skeptical of how accurate it can be measuring current as frequencies rise above the typical 60 cycle range, but i'm convinced that it showed a significant stator current difference when comparing a shunt unit to the series SH-775 .
I have had an outboard voltmeter installed from my first adventures in charging fun, so weird stuff happens when the stator starts failing. Voltage might become obviously low or erratic or it might just level off at 13.5 no matter what rpm you're at- my last failed stator hung on like this, still producing enough to keep bike running ok for another 30 hours or so. Removal of stator showed it bbq'ed in one section.

EDIT: The 400Hz spec was on voltage not current measurements.

I looked on Amazon for some other clamp on meters to see what the AC amp spec was. I have not seen another yet other than this $14.95 7 function model from HF that lists a frequency response out to 400 hz (3600 RPM with 18 pole stator).

http://www.harborfreight.com/clamp-on-digital-multimeter-95683.html

Amazon lists the HF meter for about $35 and not eligible for Amazon Prime.

http://www.amazon.com/CenTech-Clamp-On-Digital-Multimeter/dp/B004SC2FZ4

The first line of defense is to have a voltmeter and basically do a mental Quick Tests as you start up your bike and give it a rev to warm up. The erratic voltage reading are always most likely going to be due to the stator, but you can confirm that without pulling the stator using the clamp on AC amp meter.

[bwringer]

So if I understand all this correctly, it doesn't really matter much whether the clamp meter gives "true" laboratory-accurate current readings.

As long as you're reasonably consistent with the test RPM, it will quickly tell you whether one or more legs of the stator are failing with load and heat. Plus, you can do the test without disconnecting anything in many cases, as long as you can separate the wires enough to sneak the clamp in.

I like it!

[posplayr]

So if I understand all this correctly, it doesn't really matter much whether the clamp meter gives "true" laboratory-accurate current readings.

As long as you're reasonably consistent with the test RPM, it will quickly tell you whether one or more legs of the stator are failing with load and heat. Plus, you can do the test without disconnecting anything in many cases, as long as you can separate the wires enough to sneak the clamp in.

I like it!

That is basically it.
DIT: OMIT teh part in quotes. I read the spec incorrectly. It should still work regardless.

But even the cheap HF meter referenced has specified accuracy out to 3600 RPM which if you look is about where the maximum current occurs anyway(see figure in first post).

There is reasonable accuracy and any lack of accuracy doesn't really matter for diagnosis of the failed stator.

Because the current tends to saturate at the maximum output anyway (above 3500 rpm), you are in fact less sensitive to variation of your test RPM. In contrast Open Circuit (OC) voltage measurements are going to be proportional to RPM and so differences in tests RPM propagate directly to a perceived voltage imbalance. It is harder to get a wrong answer with the current clamp.

The revised Phase B tests for OC leg to ground voltage tests are still probably useful as you should see zero and even 5-10 VAC on the ground will indicate a problem before you see an imbalance currents. We have had at least a couple of people comeback saying they read 40 VAC on leg to ground with no apparent issue. I don't have any explanation as to how they did that. Maybe they did the testing with the stator connected to the R/R. That would probably explain it. The ground is then through the R/R.

[posplayr]

I edited the posts in this thread. Turns out the 400Hz spec was on voltage measurements not AC current. None of the conclusion change though.