Aftermarket magneto slip rings

Magnetoman

Well Known and Active Forum User
VOC Member
If you go to the following link you will find a new post I just uploaded diagnosing in detail the cause of the failure of an aftermarket slip ring:

http://www.britbike.com/forums/ubbthreads.php?ubb=showflat&Number=507338#Post507338

As I write there, although the results of that dissection and analysis apply only to that one specific slip ring, the implications for users of aftermarket electrical components are much broader. With most mechanical components on a motorcycle it's clear whether or not an aftermarket product works. Either the holes line up properly (or the bolt has the right pitch, or the fuel tank fits, or...), or they don't. There can be exceptions, like a badly heat-treated camshaft, but largely the average mechanic can tell when they install an aftermarket product whether it has the quality they expected.

Aftermarket electrical products are a different story, since they require specialized instruments to know whether claims about them are valid or not. This can be seen in the above post about slip rings. Without a particular highly specialized General Radio Megohm Bridge that measures to 1000 TOhm (a thousand, million, million ohms) I couldn't have made some of the measurements I made that were necessary to determine the cause of failure of that slip ring. As I say at the end of that post, the failure mode of this slip ring now has me investigating other ways for stress-testing slip rings that could be used by people who don't have the necessary specialized electrical test instruments.
 

Ken Targett

Active Forum User
Non-VOC Member
I think that specialised equipment is needed to test slip rings (and pick-ups), but I'm not sure that the 1000 Tohm meter mentioned above, or the 20 Gohm meter or 5 Tohm meters mentioned in the Britbike link, would be up to the job of detecting a failed component when one doesn't know in advance that it has failed. From what I gathered, those three instruments are measuring at 1 kV, 2.5 kV and 500 V, respectively, but sometimes failures are only apparent when the component is subjected to the higher voltages to which it is subjected in normal use. Also, testing at elevated temperature is sometimes needed to detect a failed component.

The instrument I use has a spark generator that produces a 50 Hz spark across a 3-point test gap usually set to 5.5 mm (so about 8 kV), and I compare the driving currents required by the spark generator to produce consistent sparks with and without the component connected in parallel with spark gap. A good component has a negligible effect on the required driving current. The extra current required is a good indication of how leaky the component is. The test is done not only with the component cold, but also at elevated temperature, usually 50 C. More information here for anybody interested http://brightsparkmagnetos.com/workshop/PickUpSlipRingTester/PickUpSlipRingTester.htm

Ken.
 

Magnetoman

Well Known and Active Forum User
VOC Member
For convenience, I've simply cut/pasted information on this I wrote on BritBike Forum:

Originally Posted By: Ken Tee
The instrument I use has a spark generator that produces a 50 Hz spark across a 3-point test gap ... the component connected in parallel with spark gap. A good component has a negligible effect on the required driving current. The extra current required is a good indication of how leaky the component is.


Both the Merc-o-tronic and Eisemann testers I discussed in my Bosch magneto restoration thread directly measure the primary current required for the secondary to jump an 8 kV 3-point test gap. Also, the normal way I use both instruments has the slip ring as part of the circuit. However, the point is, the aftermarket slip ring I diagnosed at:

http://www.britbike.com/forums/ubbthreads.php?ubb=showflat&Number=507338#Post507338

was fine when new but failed after ~2000 miles of operation. Because of the very high resistance of the resin that I measured (TOhms) this slip ring almost certainly had negligible leakage when new and would have easily passed such a spark gap test. However, it would have been quite incorrect to conclude it was a "good component" based on a spark gap test lasting only a few minutes because it would require running the tester for over 60 hours at 50 Hz for the slip ring to accumulate the internal degradation caused by 2000 miles of usage.

Originally Posted By: Ken Tee
...I'm not sure that the 1000 Tohm meter mentioned above, or the 20 Gohm meter or 5 Tohm meters mentioned in the other thread, would be up to the job of detecting a failed component when one doesn't know in advance that it has failed.


This is precisely the point. For reasons I have discussed in detail neither a resistance meter nor a spark tester is appropriate for determining in advance the likelihood an aftermarket component like a slip ring will fail in only a few thousand miles.

Further, as I wrote about the slip ring, "...the failure mode of this slip ring has resulted in me now investigating other ways for stress-testing slip rings that could be used by people who don't have the specialized instruments that I have available to me." Rather than accepting assurances that they just bought a "good component," I am looking into ways to provide people with a test they can do themselves to know whether or not this is the case.
 
Last edited:

Ken Targett

Active Forum User
Non-VOC Member
I totally agree with just about everything you've just said, Magnetoman.

As it happens, when I'm doing quality control checks on the slip rings and pick-ups that we sell, I choose a free weekend and leave the component on the slip ring/pick-up tester all weekend, which as it happens is about 60 hours.

The problem with using just a coil tester like the Merc-o-tronic or Eisemann to test an armature's slip ring is that it's no good if the armature windings have also failed. Customers who'd sent us their mag for servicing/repair would find it annoying if, after they'd agreed to a rewind costing £X, we then went back to them after the rewind to say that the slip ring had failed and was going to cost another £Y to replace. Admittedly, it's not always possible to diagnose everything that's wrong with a magneto from an initial assessment, but we try as much as possible to avoid hikes in the initially-estimated cost part way through the job.

Good luck, Magnetoman, in your quest to find some testing method for predicting future failure of a slip ring without using specialised equipment. Must admit I can't think of any way of doing it. In the meantime, people are going to need to rely on traders' reputations and what they have to say about their wares. A good thing about forums like this is that if somebody is dissatisfied with a magneto spare they've actually bought, or service they've received, they can tell the whole world about it.
 

Magnetoman

Well Known and Active Forum User
VOC Member
Again, cutting and pasting from BritBike Forum:

Originally Posted By: Ken Tee
I choose a free weekend and leave the component on the slip ring/pick-up tester all weekend, which as it happens is about 60 hours.


I can conduct a test on only a few capacitors and have high confidence the result is representative of all such capacitors because it is possible to rely on the manufacturers' own tests and specifications. The thing capacitor manufacturers don't normally test for is longevity in the presence of oil vapor, so my tests are directed toward that while also subjecting them to magneto-level current pulses. In contrast, a specialty product like a slip ring produced in very small batches by small companies without the reputation or quality control of a major electronics manufacturer is different. There conducing only one test tells you only about that sole component you have tested.

As I wrote in my analysis of the slip ring that failed after only 2000 miles ("60 hours"), those results only applied to that one specific slip ring.

http://www.britbike.com/forums/ubbthreads.php?ubb=showflat&Number=507338#Post507338

Before anyone could say a slip ring from a given batch was a "good component" a statistically significant number of them would have to be tested to failure or to at least the equivalent of 20,000 miles. This is what major electronics manufacturers do before selling their products. And, even then, controls would have to be in place at the slip ring manufacturer to ensure batch-to-batch repeatability.

Originally Posted By: Ken Tee
Good luck, Magnetoman, in your quest to find some testing method for predicting future failure of a slip ring without using specialised equipment. Must admit I can't think of any way of doing it. In the meantime, people are going to need to rely on traders' reputations and what they have to say about their wares. A good thing about forums like this is that if somebody is dissatisfied with a magneto spare they've actually bought, or service they've received, they can tell the whole world about it.


For a number of years one well-known supplier sold replacement condensers with the claim "Modern substitute, very high specification, zero failure." Despite this claim, they failed in service. Still, it took a number of years before enough motorcyclists complained to the supplier about failures for the supplier to cease selling them. In all that time I never saw a single complaint registered on a Forum like this, nor any on-line comments questioning that trader's reputation. This is not surprising given how uncertain most motorcyclists are about electricity.

Unfortunately, the long, checkered history of aftermarket electrical components is why it is essential to follow Ronald Reagan's advice and "Trust, but Verify." Trust, to spend the time and money buying a particular aftermarket component in the first place, but Verify to ensure it actually performs as the supplier says it does. Even if 50% of the slip rings from a given batch are bad, it only matters to the buyer if they got one of the good ones. So, the seller needs to do relevant tests on a statistically significant number of components before making claims about them, but the buyer only needs to test the one component they bought.
 
Top