1978 GS 550 Draining battery Wont start

[posplayr]

Not sure what numbers youre looking for on a "quicktest"

The quick test in in a link in my signature. There are 6 numbers.

[bkostiuk]

Key off: 12.69
Key on: 12.01
at idle: 12.25
at 2500RPM:12.47
at 5k RPM: 12.65

[JTGS850GL]

Key off: 12.69
Key on: 12.01
at idle: 12.25
at 2500RPM:12.47
at 5k RPM: 12.65

Stator resistance was .9 ohms on all 3 coils
75-85v when at 6k rpm
No continuity either way on any wires of voltage regulator.

From what you've measured so far, I'd say your R/R is more suspect then the stator. The stator resistance measurements and output are in the good range. Make sure that each leg of the stator is open to ground. If that checks out then it's the R/R and/or associated wiring that's most likely the problem.

[bkostiuk]

Infinite resistance between stator and any of the 3 legs.

[bccap]

Before you go off and buy a new RR clean up the connections and run the RR ground wire to the battery and do the quick test again. If you do get a new RR ground it to the battery. Also wait til posplayr replies, he knows this electric sh!t really well.....

[JTGS850GL]

Infinite resistance between stator and any of the 3 legs.

Good, that's what you're looking for. Like stated above... Look at each connection on the R/R and stator and verify good contact. Make sure the grounds are good. If all checks out then I feel a new R/R is in your future.

If the R/R is bad then you may consider this a good time to upgrade to the shindengen sh775 series regulator.

[posplayr]

From your test results you are charging some but very little. Stator tests can only confirm a fault not confirm if good. Test the leg to ground at 5k . This is the best stator test as it is one of the best checks for shorts to ground. You should see zero ac at 5k.

Your regulator tests are at least showing the blocking diodes are blocking did you test both ways?

I'm figuring stator but it is a guess.

[JTGS850GL]

From your test results you are charging some but very little. Stator tests can only confirm a fault not confirm if good. Test the leg to ground at 5k . This is the best stator test as it is one of the best checks for shorts to ground. You should see zero ac at 5k.

Your regulator tests are at least showing the blocking diodes are blocking did you test both ways?

I'm figuring stator but it is a guess.

Just curious as to why you're leaning toward a stator issue with all the static measurements in the good range and a solid 75-85v output at +5K RPMs? Admit that an AC measurement to ground would be an additional verification of insulation integrity, but I would think that the next likely culprit would be the R/R or subsequent wiring.

Also curious as to why no one has built a simple resistive load tester for the stator as a stand alone test of current capacity?

[posplayr]

Just curious as to why you're leaning toward a stator issue with all the static measurements in the good range and a solid 75-85v output at +5K RPMs? Admit that an AC measurement to ground would be an additional verification of insulation integrity, but I would think that the next likely culprit would be the R/R or subsequent wiring.

Also curious as to why no one has built a simple resistive load tester for the stator as a stand alone test of current capacity?

Go to my link gs charging health and find the link to revised phase b stator tests it is there.

[JTGS850GL]

I see that you've included a basic resistive star configured load test. That would conclusively rule out any stator issues currently. The test can not determine future longevity, but would rule out the stator as a currently problematic. I understand that the loaded coil arrangement test would be more conclusive, but given the results obtained thus far I would say that the available data still leans more toward post stator issues.

Just wondering why, in this case, you would lean toward a stator related problem v/s an R/R or wiring issue? From the tests currently performed I would still lean toward further investigating the R/R or wiring. This is based on all of the no load tests results provided thus far. All are within established parameters while the R/R tests conducted were very limited at best. No information on instrument setup was provided for the R/R tests. Was the DVM even set in diode check mode when conducting the tests? Even if all test were performed correctly, there is still no reliable static method of verifying the regulation portion of the R/R.

Please understand that I'm not trying to argue a point for the sake of argument... I'm just trying to understand the logic behind the conclusion so that I can better understand what I may be missing here. Are you leaning toward the stator because of historical failure rates?

[posplayr]

I see that you've included a basic resistive star configured load test. That would conclusively rule out any stator issues currently. The test can not determine future longevity, but would rule out the stator as a currently problematic. I understand that the loaded coil arrangement test would be more conclusive, but given the results obtained thus far I would say that the available data still leans more toward post stator issues.

Just wondering why, in this case, you would lean toward a stator related problem v/s an R/R or wiring issue? From the tests currently performed I would still lean toward further investigating the R/R or wiring. This is based on all of the no load tests results provided thus far. All are within established parameters while the R/R tests conducted were very limited at best. No information on instrument setup was provided for the R/R tests. Was the DVM even set in diode check mode when conducting the tests? Even if all test were performed correctly, there is still no reliable static method of verifying the regulation portion of the R/R.

Please understand that I'm not trying to argue a point for the sake of argument... I'm just trying to understand the logic behind the conclusion so that I can better understand what I may be missing here. Are you leaning toward the stator because of historical failure rates?

You did not read the whole thing apparently. There should be ample information to conclude that the testing of a stator amount to the testing of the insulation. Insulation tests can not be done with a ohm meter. The standards are based on voltage breakdown.

So when I say a conventional stator test (non - MEGGER) can only unambiguously proves a fault, it is because it those tests do not fully test the stator to the level required in operational use, then concluding that the stator will pass in operational use leaves a large gap in performance and therefore an apparent ambiguity in the results.

The stator test is only valid is it fails, as if it fails at a low level you can be assured it will fail at a higher level. I can state this in the following RULES for testing stators.


FIRST LAW OF STATOR TESTS:
ANY POSITIVE RESULT of a STATOR TEST HAS AN AMBIGUITY PROPORTIONAL TO THE UNTESTED OPERATIONAL STRESS.

SECOND LAW OF STATOR TESTS:
ANY NEGATIVE RESULT of a STATOR TEST IS A NECESSARY AND SUFFICIENT CONDITION TO DECLARE A STATOR BAD.

[posplayr]

FIRST AND SECOND LAWS of Stator Testing

I started a new thread here.

http://www.thegsresources.com/_forum/showthread.php?t=223164

[JTGS850GL]

I read it quite thoroughly and fully understand the tests. My question is how conclusions are drawn from the tests that were conducted in this case.

Please bare with me. Let me see if I understand you correctly. Weather a stator fails any test (positive failure mode verified) or completes only a portion of the full test battery (albeit without a failure IE: test battery incomplete), then you always suspect the component as being bad. This is very true when attempting to identify a component or system that is deemed to be good only after all tests have been completed. In my previous life as a RF test engineer in the cellular communications industry I would agree with you whole heartedly. Any device that hasn't completed all the required tests should not be considered as fully functional. This also assumes that the test battery is fully comprehensive as well. Lord knows I've seen enough incomplete test systems implemented over the years. With that said, this may not be the best method of identifying which component is likely the point of failure on a know dysfunctional system.

In this case I would lean more toward a probability method of diagnosis. Since no test results indicated a conclusive failure and all test that were conducted indicate a positive result, I would say that probability of failure would lean more toward the component/s that have not been tested or yielded no positive test results thus far. In this case the R/R and subsequent wiring are still a big unknown. Without the ability of the end user to complete all tests required on the stator the only option you leave would be to replace the stator in all cases with a known good component first before preceding to other options. From a linear stand point this would make sense since the stator is the furthest up stream component in the system.

The reality is that, without proper diagnostic tools, there would be no definitive method of identifying the stator as being fully functional. In fact there may be multiple failure modes present in this case which would complicate diagnostics even further. I'm simply saying that this may be a case where (if you have to shotgun component replacement) I'd start with components further down stream of the stator. Could the stator still be defective? Yes. If I was going to throw money at the problem I'd still start with the R/R AFTER I checked all associated wiring. Replacement cost is lower for the R/R and a possible upgrade (sh755) could augment system functionality. Future wiring issues could also be addressed simultaneously with the R/R replacement.

Again, this is not intended as an argument. I'm simply presenting another option of identifying the most likely component given the limited amount of test results presented.

[posplayr]

I read it quite thoroughly and fully understand the tests. My question is how conclusions are drawn from the tests that were conducted in this case.

Please bare with me. Let me see if I understand you correctly. Weather a stator fails any test (positive failure mode verified) or completes only a portion of the full test battery (albeit without a failure IE: test battery incomplete), then you always suspect the component as being bad. This is very true when attempting to identify a component or system that is deemed to be good only after all tests have been completed. In my previous life as a RF test engineer in the cellular communications industry I would agree with you whole heartedly. Any device that hasn't completed all the required tests should not be considered as fully functional. This also assumes that the test battery is fully comprehensive as well. Lord knows I've seen enough incomplete test systems implemented over the years. With that said, this may not be the best method of identifying which component is likely the point of failure on a know dysfunctional system.

In this case I would lean more toward a probability method of diagnosis. Since no test results indicated a conclusive failure and all test that were conducted indicate a positive result, I would say that probability of failure would lean more toward the component/s that have not been tested or yielded no positive test results thus far. In this case the R/R and subsequent wiring are still a big unknown. Without the ability of the end user to complete all tests required on the stator the only option you leave would be to replace the stator in all cases with a known good component first before preceding to other options. From a linear stand point this would make sense since the stator is the furthest up stream component in the system.

The reality is that, without proper diagnostic tools, there would be no definitive method of identifying the stator as being fully functional. In fact there may be multiple failure modes present in this case which would complicate diagnostics even further. I'm simply saying that this may be a case where (if you have to shotgun component replacement) I'd start with components further down stream of the stator. Could the stator still be defective? Yes. If I was going to throw money at the problem I'd still start with the R/R AFTER I checked all associated wiring. Replacement cost is lower for the R/R and a possible upgrade (sh755) could augment system functionality. Future wiring issues could also be addressed simultaneously with the R/R replacement.

Again, this is not intended as an argument. I'm simply presenting another option of identifying the most likely component given the limited amount of test results presented.

I re read your post again, and you are misinterpreting. It is only bad if it tests bad. It is not good if it tests good. So any good test result is ambiguous and there is no conclusion to be drawn unless you understand the extent of the test. Most people can not determine the extent of the stress of the test so I have not done any further delineation of the ambiguous result. Simply stated you dont know enough to act on it. That is as simple as I can say it. Here it is mathematically.

Total probability = P(F/F) + P(F/G) + P(G/F) + P(G/G)

P(F/F)=1-nearly zero any failure at any stress level up to maximum stress implies failure at any stress level above that.
P(F/G)=nearly zero; measurement error and is considered small when measuring 65-80VAC.
P(G/F)=1-F(sufficiency of the stressing condition)
P(G/G)=F(sufficiency of the stressing condition)

Observations:

If you complete the least stressing test and get a failure, then the presumption is that any more stressing test will also fail resulting in a correct diagnosis of a true failure with any level of stress (less than the full operational stress).

If you complete any test that is less stressing than full operational stress and you pass, there is a finite probability that increasing the stress of the test will fail resulting in a false positive.

The laws directly follow from these observations.


I never said testing the stator was an optimal approach to diagnostics. In fact I have said the opposite, due to the shot in the dark stator testing of getting a failure to confirm a BAD diagnosis, the other outcome is ambiguous and there is no information.

I have said the Quick TEST is preferable to the Stator pages, and the reason is that the Quick Test is capable of generating a measurement vector (6 measurements) with a set of potential patterns that can be used to deduce proper operations from a variety of failures. There are only 6 numbers so it is not possible to separate all failures from each other, but it is a full test of the system and in that regard a i=iterative process fo fix and test will eventually result in a working system. Given the easy fo performing the test, there is little penalty fro retest and there is only minimal requirement for a volt meter which should not be considered a burden.

I don't disagree with anything you said except attributing the "use of stator tests" as a primary mode of diagnostics to me. I never said or did anything to imply that. By a process of elimination, (swapping out stator, or R/R) the charging system failures can also be diagnosed. Yes and yes of course. The optimal choice is that which achieves the fastest repair time. If swapping something out is fast and has a finite probability of fixing the problem, it may very well be the optimal solution to diagnose and fix the problem. I dont think I have said anything to contradict that as that is the fundamental premise from which I view the problem.

The issue of how long it takes to perform a diagnostic step and the optimality of taking that step is fundamental and it is the basis for the name of "Quick test".

[JTGS850GL]

Clarification understood.