Feynman ex Machina
July 11th, 2025That Little White Thing Behind the Grass is my House
The Kubota mower saga continues. The mower which was supposed to be my deliverer has become my captor, forcing me to spend hours and hours working on it. Yesterday I had to come up with a new strategy to fix it.
Clearly, this means I have to discuss Richard Feynman.
Being smart is a good thing, obviously, but it has its drawbacks. It can actually cause you to behave stupidly in practice. If you go through life figuring things out effortlessly or having answers just come to you, you can find yourself floundering when a puzzle is actually challenging, because you haven’t done what less-intelligent people do when confronted with hard questions. You haven’t come up with logical strategies to solving problems. As a result, you may find yourself struggling with problems people with less brains cope with more easily. You sit there waiting for answers to come, cycling through ideas that have already failed, like a bear pawing at a combination lock.
Richard Feynman claimed his IQ was 125. And Oprah claimed she lost like 150 pounds in about a month without drugs. Feynman had an impish sense of humor, so I guess this was one of his little jokes, intended to upset less-intelligent people who knew their IQ’s were up in the 180 range and make them wonder how Mr. 125 had left them in the dust.
I don’t think his IQ was 125, but I think that if it had been, he would have punched above his weight because of his approach to thinking.
I read a story he wrote about ants. He saw ants marching around in a dorm room, and he started asking himself why they did certain things. Using bait and some other handy objects, he was able to do several experiments and uncover some pretty remarkable facts about ants. He was systematic. He developed logical approaches. He didn’t sit back and think, “I’m a genius, so I’ll know the answer in a minute regardless of what I do.”
I thought about that story yesterday after a most miserable day working on my new mower in the heat and roasting sun. It was starting intermittently when I got it, and then it just plain quit, and I had to start hot-wiring it.
When I looked for answers, I decided the smart thing was to take care of the easiest and/or most likely and/or cheapest fixes instead of getting into a lot of diagnostic drudgery. If they worked, great. If not, I would have new parts in my mower which couldn’t hurt anything and might help it stay running longer without problems.
This kind of approach only works when you get lucky, which usually happens for me, but in this case, didn’t.
Having suffered considerably in the process of replacing two safety switches that were likely to have failed but actually hadn’t, I decided to change course and take the Feynman approach. I went to AI sites, I looked at the workshop manual, and I tried to make deductions and formulate a plan.
I came up with 5 problems that could prevent a mower from cranking. The seat switch had been bypassed, and the crude, bubbastic bypass connection could be failing. A safety switch on the brake, the PTO lever, or one of the mower control levers could be bad. The starter or solenoid (one indivisible package) could be bad. The starter relay could need replacing. Finally, the ignition switch could be no good.
It was also possible the mower’s electronic controller needed replacing, but that was too unlikely to consider at this point.
By the time I made the list, the switches on the PTO lever and brake had already been replaced needlessly, and I knew the starter and solenoid were fine, so I moved on. I tried to come up with a smart sequence of things to look at.
The seat switch splice had to be my first stop because it could get me going in 5 minutes. It seemed likely to be the problem because the mower had started once after I moved the seat. It was also an extremely simple fix. Find the splice and redo it with a wire nut. Tools not required.
Number two: the starter relay. This was not as likely to be the culprit as the safety switches, but checking it with an ohmmeter was a 10-minute job requiring the ohmmeter, a wrench, and a screwdriver.
Number three: the safety switches. The remaining switches were not nearly as hard to replace as the ones I had already worked on, but they were harder to deal with than the relay, so I would leave them for later. Testing them would be simple. They had to be open for the mower to run, so I can just disconnect them.
Number four: the ignition switch. Kubota didn’t make a point of creating handy test points for people with bad ignition switches, and getting the switch out to test it is unpleasant. I could test it by opening a connector on the wiring harness to expose a couple of conductors. That would be more pleasant than removing the switch, but still unpleasant. The switch worked fine in every position except the start position, so I figured it was probably okay.
Today I identified the starter solenoid by the colors of the wires going into it, and it took me two minutes or so to get it out. The manual lists the proper resistances between the terminals, and the relay failed one test. I went to Ebay and ordered three new relays (the mower uses the same kind of relay for three things), and they are on the way. They’re $10 each, and if one has failed, the others could be on the way out.
Will the new relay get my mower running? No idea, but it should. I can test my theory by moving one of the other relays to the starter position. One relay only controls a horn that goes off when the mower overheats, so I can do without that for a few days.
I should have switched the relays after testing the bad one, but what can I tell you? It’s hot outside. Feynman would be ashamed of me.
If the relays don’t help me, I have to test the safety and ignition switches, and after that, I have to consider the possibility that a $332 controller has bitten the dust. I think I tested the relay correctly, though, and the mower can’t start if it’s bad, so I have every reason to think I will not need a controller.
I caused myself a lot of very unpleasant and unnecessary work by guessing at the mower’s ailment. If Feynman were here, he would have gotten where I am in two days. Nobody could do it faster except someone who is familiar with Kubotas. Two days is how long it would take to understand the starting system, including an hour or two to come up with a plan.
I should try to take the Feynman approach to problems from now on, during what little remains of my life. Prayer first. Then research and logic. I can’t get back all the hours I’ve spent nullifying my intelligence by hoping answers fall out of the sky, but I should be able to benefit to some limited degree.
July 13th, 2025 at 8:01 PM
Electrical problems seem to be the most difficult to diagnose.
Not that I’ve had many to work on personally, but my brother has a 1976 Suzuki GT 550, and I’ve watched him spend days trying to figure them out.
Not something I have the patience for.