December 14th, 2009
Today I set out to repair my tractor. For some reason it wouldn’t start. I’d engage the starter and I would hear the starter motor spin, but the engine wouldn’t turn over. Why wasn’t the motor engaging the engine? It probably had something to do with the very low temperatures of the last few days, but what?
Aha! I remembered a little tidbit from one of the Feynman lectures on physics: rubber, unlike most other materials, contracts when heated and expands when cooled. The fan belt is made of rubber, I think, or something very similar to it. Ergo, the fan belt has expanded in this cold and is now loose, so the starter motor is spinning but not moving the fan belt. All I had to do was tighten the fan belt.
Just as I finished tightening the fan belt, I noticed something I had overlooked: the starter motor on the tractor, unlike that on most cars, does not drive the engine via the fan belt. It engages a geared flywheel. How very stupid of me! OK, so it wasn’t the fan belt.
So it must be that the solenoid that engages the starter motor with the teeth of the flywheel is stuck. That’s a common problem on old Volkswagens, and when that happens, all you have to do is hit the starter motor with a hammer when it’s trying to turn over. So I got Kathy to engage the starter while I hit it (not too hard!) with a hammer. It didn’t work.
At this point I decided that I needed to get analytical, so I pulled out my voltmeter. I started the battery charging and waited a few hours. Then I checked the battery voltage: it was 7.20 volts, a high voltage for a 6-volt lead-acid battery but it merely reflected the influence of the battery charger. I hooked up the voltmeter to the input connector on the starter motor and engaged it; the voltage on the starter motor was a miserable 0.91 volts. Aha! That’s the problem! The starter is not getting enough voltage. Could it be the connections leading up to the starter?
I disconnected the starter and checked the voltage again: it was a solid 6.36 volts, the same as the output of the battery. No, we weren’t losing voltage in the switches -- however, if the switch connections were weak, we could be losing a lot of voltage at the bad connections. However, if that were the case, we’d be pumping a lot of current through those weak connections, and they’d get real hot real fast. That obviously wasn’t the case. That left only one possibility: the starter motor had some kind of short in it. It probably hadn’t been torn down in decades, and the brushes could well have worn down to stubs and were now shorting.
So I pulled the starter motor off the engine, took it into my shop, and completely disassembled it. The brushes looked fine; perhaps a little dirty, but certainly not worn oddly. The armature looked solid, so I just cleaned everything as best I could. Then I reassembled it all and mounted it back on the tractor. This time, when I engaged the starter, it would surely work!
It didn’t.
That was when I finally realized that the entire problem lay in the battery. It had died and was no longer holding enough charge to power the starter motor. All that work I did rebuilding the starter motor had been a complete waste of time.
I share the tractor with a neighbor, so I called him and related the entire story. He was almost horrified that I had torn down the starter motor for no good reason.
"Look on the bright side", I told him. "At least I’m not your doctor."