Classic Harley Tech: Bosch ZEV Magneto Reassembly, Remagnetizing & Extended Stress Tests (Continued):

As with all the articles in this series on rebuilding a Bosch ZEV magneto, this one is reposted with permission of the author, who requested we not share his name.

Long-Term Tester

The reason for conducting an extended test is that, assuming a motorcycle is geared to go ~40 mph at the engine speed corresponding to my test speed for the magneto, a 12-hour test is equivalent to it covering 500 miles. Any teething problems should be revealed in this length of time, but it is not so long that it would cause significant wear on the “new” magneto. In the case of this Bosch ZEV magneto for the Harley-Davidson, I planned to extend the test to 24 hours.

My lathe has continuously variable speeds from 40-2000 rpm so I could use it to spin the magneto at whatever speed I want to within this range for further tests. However, since I run magnetos for at least 12 hours after rebuilding them I don’t want to unnecessarily subject my lathe to that much use. So, some time ago I built a dedicated magneto tester using a reversible 1/2 h.p. motor and a universal mounting base that lets me test every type of platform- and flange-mount magneto I’ve yet come across. With my usual pulley the motor spins the magneto at 2000 rpm (4000 rpm engine). However, after finding with the Strobotac that the points on this magneto float at that speed, I had to fabricate a larger pulley in order to slow it down. Doing this was straightforward, but it ate up valuable time as the deadline for shipping it back approached.

I wanted the new pulley to be ~4″ dia. to cut the speed by a third, but didn’t have any Al bar that large at hand. So, I bought a 4″ pulley from the hardware store, machined an Al rod a thou. oversize to press fit in the pulley’s bore (and also held with the set screw), and then used my lathe to bore the necessary 1:10 taper for mounting it on the magneto. With this new pulley the magneto now spins at 1400 rpm (2800 rpm engine), which should be close to the upper limit it will experience on this motorcycle. Checking with the Strobotac showed that the points no longer bounced at this lower speed.

The first thing I found when I started the actual test was that the spark from this magneto was so hot that within a couple of minutes it began melting the plastic insulator on the 6-gap board at the left side of the tester. I got this Merc-o-tronic board on eBay recently and installed it in place of the one I had made myself (the fact I connected it to leads 5&6 instead of 1&2 is irrelevant).

Rather than take the time to reinstall my old gap board, instead I attached the two HT leads to nylon screws that can be seen at the right of the housing in the above photograph of the tester. I then used 0.032″-dia. stainless wires to make 5 mm gaps, corresponding to 6 kV at atmospheric pressure, and continued the test. This voltage is about 50% higher than what will be required to jump the spark plug gap in the operating engine.

At the left of the above composite photograph you can see that ~1 mm of the wire nearest the tip is glowing red hot from the heating caused by the spark current flowing through it (the bluish sheath around the tip is from the ionized plasma created by the high electric field, and the out-of-focus wire from the other lead is in the background). Since each spark lasts only a msec, it’s easy to capture an image when it is not sparking, and that is what is at the right. Here it is even easier to see that the tip of the wire is red hot. What this shows is that, even though each current pulse lasts only ~1 msec. and is separated from the next by 35 msec. (i.e. a duty cycle of less than 3%), enough current flows to maintain the tip of the stainless steel wire at ~1000 oF (i.e. the temperature where the steel glows red). Note, though, that a spark plug electrode has both a larger diameter and a higher thermal conductivity than the stainless steel wire, so it would not get nearly as hot as the wire I used for these tests.

Even more dramatically than the glowing wire, if I slip a piece of paper between the electrodes it immediately bursts into flame (it did not touch the hot electrode; it is the spark itself that ignited the paper). Since the sparks from this magneto so easily set fire to paper, I can be reasonably confident they will ignite the mist of gasoline in the combustion chamber.

Elevated Temperature Test

After running the magneto on this tester for 18 hours, I wrapped it in heating tape and heated it to ~50 oC (122 oF) using a Variac, with a thermocouple to monitor the temperature. I ran it another 6 hours at that elevated temperature and it continued to spark reliably. Earlier, in the interest of time, I had skipped testing the armature by itself at elevated temperature, and this 6-hour test of the full magneto vindicated having taken that shortcut. However, had the magneto failed this elevated temperature test due to a faulty armature I would have had no choice but to disassemble it and find the time to wind a new coil myself. I took the next photograph after completing this test and already starting to remove the heating tape. Send questions or comments to

This entry was posted in Harley Tech, Uncategorized and tagged , . Bookmark the permalink.