Ignition condenser/capacitor post mortem

I received two failed capacitors (I use that term, not the older term "condenser") from Scott Westgate (thanks!) and dug into them today. As a result, I think I know why these things are failing. Anyway, here is the story...

I examined three capacitors, the two failed ones from Scott and a third, unused one of mine, as a control. These are numbered 1 to 3; mine is no. 3 (first picture). Nos. 1 and 2 were distinctly different, apparently from different manufacturers. Mine was marked "Made in England," so it could have been a Lucas unit, but I'm not sure.

Electrical Tests

First, I tested all for open or short circuits. All were OK. Next I checked the capacitance, and all seemed OK this way too, with capacitances between O.20 and 0.22 microfarads (the spec is 0.18 to 0.25). Surprisingly, they seemed to be working. I also checked to see if there was any breakdown. Since the ignition system subjects them to about 250 volts, peak, it seemed possible that the high voltage might have blown a hole in the insulation, and this would not be evident at low-voltage measurement but would allow arcing at high voltages. I don't have a high-voltage power supply to test this, but I did cobble together enough to get 175V. They all held up OK to this level, at least.

Physical Examination

Next I disassembled them. I was looking for a sign of failure, such as arcing across the capacitor or through the dielectric. Here's where things got interesting.

I started with no. 1. I ground off the edge of the case, so it could come apart (second picture). Inside was a capsule (the capacitor itself) and a corrugated nickel-plated contact. The other contact was a brass washer to which the wire was soldered. The capsule, as I'll call it, consists of two layers of foil with a thin mylar layer between the metal layers. It's all rolled up into a cylinder, with one metal layer sticking out one end and the other layer sticking out the other end. Electrical contact is made to the ends, by simply squashing it all together. The capsule is smaller than the can in which it is mounted, so it conceivably could wiggle around a bit in response to shock and vibration. Not good.

The whole thing is a stunningly cheap, unreliable design. There is nothing to keep some force on the capsule contacts; without this, there is no guarantee that the electrical contacts remain good, especially considering the fact that the mylar will expand and contract somewhat with temperature changes. There should some kind of spring, to keep the pressure on the contacts, but there is absolutely nothing, not even anything the least bit springy. The lack of any support for the capsule is also troubling, as it allows it to move around if it comes loose, which will also affect the electrical contacts.

This is all especially important, since the capacitor carries a surprisingly high current. I estimate it at about 2.5 amps, peak. This means that even a small additional resistance is likely to affect the ignition pulse and waveform, with serious affects on the operation of the car.

I started by unwrapping the capsule, so look for signs of insulation breakdown (third picture). There were none. I then took a look into the can and examined the corrugated washer, and saw some clear indications of arcing at the contact points between the washer and the can. This is shown in the fourth picture, although it's hard to see in the photo, especially inside the can. It is quite clear in real life, though.

No. 3, the presumed Lucas cap, was a little better quality, but not much (pictures 5 and 6). There were no marks inside it, which is no surprise, since it was unused. It had no corrugated washer, so the can-end electrical contact was simply to the body of the can, and the other contact was a brass piece that was crimped to the wire instead of being soldered. Again, the whole thing was simply squashed together with no spring of any kind to keep the pressure on the capsule. The capsule was also considerably smaller than the inside of the can.

I finally opened no. 2, and that was quite a surprise. While its design was very similar to 3, it showed clear indications of significant arcing at the can end. Not only were there marks on the can, but there were marks on the capsule as well. It had arced quite a bit (picture 7).

Conclusions

it appears to me that, because of a stunningly boneheaded design, these capacitors are failing because of a poor electrical contact at the can end of the mylar/foil capsule. Since there is nothing to maintain pressure on the capsule, eventually it starts to loosen, because of temperature variations and/or vibration, and the can end begins to arc. This creates additional resistance, and may even cause an intermittent open circuit in the capacitor, which in turn results in bad operation, burned points, and ignition failure. It is also consistent with reports that new capacitors often seem to work well for a period of time, then fail.

What to Do About It?

Frankly, I find the stunningly poor quality of these components more than a little disturbing. I view these capacitors as time bombs in the ignition system. The chance of one failing, sooner or later, is pretty high. Such sloppy manufacturing and design quality really should not be tolerated in cars that have enough potential reliability problems already.

So, what to do? Best bet, in my opinion, is a commercially available capacitor that might be a good substitute. One option that looks good to me is a "snubber" capacitor. These are designed for soaking up large voltage spikes, much like what the capacitor in an ignition system should do. They are not cheap by capacitor standards, a few bucks, but not expensive by automotive standards. I'll look around, and post again when I've found something that looks good. I don't use a conventional ignition system any more, so I'll need volunteers to test whatever I come up with.

WOW! Just image.. That cap is bolted directly to the dizzy stuck in the engine of an A series engine that has horrible harmonics-- vibrates like crazy. It has to be taking a beating. I can clearly see why they fail just rattling around in that capsule(can).That is a poor quality part.
Why couldn’t it simply be mounted off the engine say on the coil bracket? At least it would have a better chance of surviving there then on the buzzing engine. They probley moved the coil from the generator for just that reason.

Yes, I did read once that the coil was moved from the generator to the car body because vibration was damaging it. I find that quite believable.

Actually there is no reason why you couldn't mount the capacitor on the car body somewhere, as long as there's a good electrical connection to the body, and run the wire to the coil terminal that goes to the points. Might help it last somewhat longer, but the design and manufacturing quality of this part is so slipshod that I doubt it would be a full solution.

As I believe I mentioned in the last thread on condenser failures, I used to work for Sprague Electric, a former manufacturer of quality capacitors.

You did an excellent job of dissecting and documenting what you found inside the capacitors. The general construction details you found don't surprise me too much apart from the lack of an electrolyte inside the cans. I always assumed the condensers were filled much like the electrolytics we used to produce in aluminum or tantalum. The termination method with the spring washer is a bit cheap. If I remember correctly, our Mylar caps were terminated with a flame sprayed aluminum disk on each end of the coiled material.

I suppose that any number of alternate capacitors could be used since the condensers are not polarized. Mylar or polyester caps are often rated for several hundred volts. I just looked around at mouser.com and found several metalized film caps rated at .22uF and 400V. There were caps for as little as $0.70 each.

Another alternative would be to use a Bosch condenser for the air cooled VW engines. I believe those have the same ratings and they are made to mount "external" to the distributor. As you pointed out, you could connect the lead from the condenser to the coil (-) terminal and simply mount the condenser to a ground point in the engine compartment.

Thanks for all the hard work you put into this. It is greatly appreciated.

Probably not filled because with the heat it would probably leak or ooose and get on the points. I have been through 3 in the past year. That is about half the number that have failed on me the previous 43 years of my life. Putting an electrolytic cap in a cooler place and running a wire might be the only real option. Lots of that stuff is coming from taiwan and china and its just crappy, I have bought some motorcycle parts from china and its a real crap shoot as to whether its going to be any good or not.

Thanks for the hard work and keep us posted!!! I've heard that almost all ignition capacitors are made now in mexico and there reliability is terrible.
Don't know if that is fact but do know of many problems with new capacitors. I have been driving old japanese vehicles since the 70's and never had a condenser fail till recently when I encountered one that had been recently been replaced by a previous owner.
For brit vehicles I fortunatly have a life time supply of NOS.
KA.

Thanks for the very thorough investigation into this problem, Steve.

So were these parts once made reasonably well, 'back in the day'? My car was running on a single capacitor for several years before all this took place.

Regardless, this thread has certainly convinced me to upgrade to an electronic ignition.

Thanks again.
Scott

Sarastro said:

Yes, I did read once that the coil was moved from the generator to the car body because vibration was damaging it. I find that quite believable.

Actually there is no reason why you couldn't mount the capacitor on the car body somewhere, as long as there's a good electrical connection to the body, and run the wire to the coil terminal that goes to the points. Might help it last somewhat longer, but the design and manufacturing quality of this part is so slipshod that I doubt it would be a full solution.

Click to expand...

You did a great job of dissecting those things and showing there flaws.
Short term moving it out board mounting it to the coil will work but finding a good source for those that choose not to run the pertronixs or the crane unit would be the thing. Finding a replacement that fits in there maybe the challenge.

Thanks for the good work. Now I am concerned.

Jeff, you around. Looks like something you would be intrested in.

I too would like a solution and look forward to one.

Thanks for the info. I'm not at all clear on what they do for the car? what the symptoms of failure are? and why I haven't seen one on Ms Triss?

Someone will correct me if I get it wrong, but their function is two fold. .. they, along with the coil, form a tank circuit, and they bring the current into phase with the voltage at coil discharge.

The reason for the capacitor: if you have a current in an inductor, which is what the ignition coil is, and you suddenly open-circuit it, you'll generate a huge spike of voltage where you break the circuit. That's the points. Then, the resulting arc would fry the points pretty quickly.

With the capacitor, the voltage rises more slowly and peaks at a bit over 200 volts. The inductor and cap make a resonant circuit, so the current becomes AC for a few cycles and dies out fairly quickly. That's just fine, because it's what the coil, which is a transformer, needs.

So, if the capacitor open-circuits, the points die pretty quickly. While they're still alive, you don't get the voltage waveform that the system is designed for, so you don't get much of a spark at the plugs. If the cap short-circuits, you get nothing, because it shorts the points, so the current is not interrupted.

In other news, I ordered a few electronic capacitors today to try out as replacements. You have to be a little careful, since the capacitor has to handle 2-3 amps of current, and not all can do this. The ones I ordered were specifically designed for power electronic circuits, so they should handle the current, but they cost a couple bucks instead of ten cents. But that's still less than the automotive ones cost, and a little overdesign won't hurt, I think. They should be here in a week or so.

so, since I have an electronic ignition I don't have a capacitor right?

JPSmit said:

so, since I have an electronic ignition I don't have a capacitor right?

Click to expand...

That is correct.

Kim, heat can affect fluid filled electronics but the seals are generally up to the task... when the part is well made. (Look at most ignition coils for example). One of the tests we performed for the military on our electrolytic tantalum capacitors was temperature cycling to 100oC while under load/charged. We had racks of the capacitors mounted on test boards inside industrial ovens. Occasionally one would "let go" from the stress. This was disastrous as it would take out many surrounding "good" components and ruin the test board. It was a price we had to pay for military contracts.

Steve, you are a real resource to do this for those of us who can not.

Awaiting your further analysis.

Thanks, Jack. Yeah, I've been an electrogeek since childhood. I get off on this kind of thing. That and little British cars, of course....

Wow, that is fantastic information! Great root cause analysis. Will be looking forward to your experiments with the power capacitors to see what works.

Go ahead and run that old-fangled spark stuff if you must, but keep a second dizzy in the boot with a nice old-fashioned Petronix mounted in it in case you get stranded.

Glen "no points" Byrns

Now folks, I've never had a failure with a points ignition that I could'nt easily fix on the road. Points and condenser take up less space and pocket change than a spare pertronix!!!
Antique Kurt!

I'm in the points-n-cap camp, too. But as we get more and more into digital and I.C. reliance it gets harder to find reliable manufacturers of "old tyme" bits. Doug worked for one of the last QUALITY makers of "discreet" electrical and electronic parts.

I've had no issues with older NOS caps, haven't had need of the newer generation yet. I, too, used to unwrap can-style & smaller caps and such as a yout... seeing the cheezy way those new ones are constructed puts th' Fear in me!

NAPA seems to still have good (enough) componentry for now. I run a Bosch (same as the above mentioned VW) dizzy in the Alfa and in 2000 put a NAPA supplied set of points 'n condenser in it... no issues after almost 80K miles. Points have been replaced a few times but if they aren't pitted I don't see a reason to swap the cap out. It's been in there nine years now.

<shrug>