Re: Why is it so hard to test a Condenser? (Was: Condensers for Johnson / OMC)

[John]

quote VinTin:

quote jnjvan:

If you really want to thoroughly test a condenser, you need to apply real world conditions. These include the voltage and current that they see in service, as well as heat and vibration. Merc-O Tronic testers, Stevens testers, and the old magic eye testers all are capable of testing at the correct voltages and (possibly) currents. The problem with bench testers is that they don’t include the heat and vibration part. I believe that heat is a primary contributor to parts like condensers and coils that test OK on the bench, and in a quick run in the barrel, but fail three miles from the launch ramp.

I work as an electrical engineer, designing electronic controls for commercial vehicles (i.e. trucks). The primary enemies of electronic components in harsh environments are heat, vibration, humidity, and deterioration due to aging. All of these are present in old outboards.

John Van

Hi John,

Once the condenser fails under real life conditions, are you suggesting that it may still test good on a bench tester?

My understanding is that once the dielectric is compromised, the cap is bad from then on and a bench tester would pick it up.

I don’t have a capacitor checker but I often test them for short or open with a volt meter and watch the resistance rise as they charge up, usually going to the infinite.

Any advice about bringing an old capacitor that has been sitting idle for decades back to spec by applying a very slowly increasing voltage giving the cap time to ‘reform’? I know it’s a technique used on long idled audio equipment and wonder how it might apply to the condensers we find in our old outboard motors. I’ve never found an open or shorted ignition cap in my limited experience with old outboard motors.

Thanks,

Dan

Hi Dan

In my experience, most components do fail hard. That is, once failed, they stay that way, and a bench test will show the fault. In the case of old outboards, where the failure may be due to degradation over time (while sitting in the barn), a bench test may not be capable of detecting a "latent" failure. That is, the dielectric may be compromised to the point that it is ready to fail, but still passes the bench tests.

Don’t get me wrong, I see plenty of value in bench tests. In fact, degradation of the dielectric could probably be detected in most cases, with a tester that looks for leakage at high voltage.

Reforming can be effective with electrolytic capacitors, which use a layer of aluminum oxide for thier dielectric. The reforming process is probably not effective with the bipolar caps used in magnetos. Most vintage caps use a wax impregnated paper dielectric, which breaks down over time (decades). Moisture and heat both accelerate the breakdown, but it still occurs slowly at room temp in a dry environment. This is why I don’t trust NOS condensers. Modern film capacitors (the film is a plastic such a polypropylene, which replaces the waxed paper) do not have this degradation issue. Note that some of the older, pre-WWII caps used mica dielectric, which is more expensive, but less prone to degradation than paper.

While we’re on the topic, not all modern film caps are suitable for magnetos, either. The instantaneous currents in these caps, which is proportional to the dV/dT across the cap, can be very high. Not all caps can tolerate this, and some will fail in short order. For this reason, the voltage rating of the cap, by itself, is insufficient to determine suitability. Jeff Register, and others, have done some great research to find modern replacement capacitors which will work reliably in outboard magnetos.

John Van