Got to thinking about this b/c of a car-repair magazine article. The owner’s older car had a misfire. Testing suggested it was most likely ignition system related. The owner dutifully replaced the crank position sensor, cam position sensor, the coils, the spark plugs, plug wires, & timing chain. Misfire stubbornly remained, unchanged. After 6 months of not solving the misfire, owner decided to take a look at the cylinder head’s electrical ground situation. Sure enough, the cable that grounded the cylinder head to the chassis was looking marginal at best. After new cylinder head grounding cable (less than $10, new), misfire completely gone.
Anyone else want to chime in with one of their most difficult to solve car diagnosis & repair problems?
I don’t like difficult because you can make a lot more money with the easy problems. In 99 I bought a 73 Maverick that hadn’t started in two years. The negative battery cable was so old and insulation cracked that it was obvious. Seventeen dollars plus a new battery and it rolls. I put five quarts of 30 weight in it and it only smoked on startup instead of all the time. I paid a hundred for the car and sold it for three-fifty. Somebody wanted to hot rod it because it was a 73 and smog-exempt. If not for registration and towing I would have made money.
Both at work and at home it is the electrical problems that get me.
Had a '94 Celica, base model that uses the Corolla engine of the time, that had all kinds of problems with the electronically controlled transmission shifting at all the wrong times. All the wires and sensors checked out which completely stumped me. I even replaced the fluid and the shift solenoids with much newer ones from the salvage yard that I bench tested. I was defeated until I found a forum thread that discussed capacitors in the speedometer assembly that are upstream of the inputs to the ECM/PCM/ECT/whatever module, after I soldered new ones from digikey in place the problem was solved.
Replacing the speedometer would not have worked because of age-related degradation (even when unused or just on the shelf) of these types of capacitors.
My 93 Toyota pickup at around 200,000 miles began to hesitate slightly at times at highway speeds. I checked all the sensors, replaced cap, rotor, plugs, wires, coil, igniter, fuel filter, fuel pump. Adjusted valves and checked compression. All good. The only thing I couldn’t check was the ECU. Factory service manual says to check it by replacing with a known good unit. Bought a used ECU at a junk yard for $100 that fixed it.
You are probably right. My memory isn’t all that great anymore. But anyways it wasn’t a “Corolla” engine, it was a Toyota engine used in multiple models.
I think my toughest one was a P0509 on my 2002 Saturn. A P0509 was idle too high. Replacing the IAC was the most common recommendation.
When the code appeared, I noticed the idle was a bit rougher, but not enough to cause a misfire code and the gas mileage dropped a little, from around 38 to 36 mpg.
Over the next 40k miles, I tried every recommendation, none worked. I scoured the web, nothing until someone else reported the same issue on a Saturn forum. Turned out to be the intake manifold gasket. It only affected 2000-2002 single cam engines.
GM had switched from a metal backed gasket to a cloth gasket with a small bead of RTV built into it. The #1 cylinder intake port only had three bolts around it because of the location of the power steering pump in that location. So the upper right corner had 180 degrees of gasket between the bolts and it was prone to failure at this point. It actually sucked the gasket into the port at that corner. Replaced it with another genuine GM gasket and it failed about 40k later, replaced it again with a felpro metal backed gasket and had no more trouble.
My 2002 Miata with 45-50K miles developed an intermittent P0103 or P0100 code. I did all the things everyone recommended: plugs, cables, coil packs, fuel injectors, ground cables, fuel filters, all with no real change. Compression was OK, too. Finally I checked the valve clearances and there it was, tight at several exhaust valves. Once I got the clearances right it started better, ran better immediately and never threw another code.
That’s a weird problem indeed. I wouldn’t have expected a p0103 related to overly tight valve clearances. But I guess if the exhaust valves don’t close and seal completely on the intake stroke, it could allow extra air flow into the cylinders, and confuse the maf sensor circuits. It’s also unusual b/c there was no effect on the compression readings.
Mom’s 1990 Mazda Protoge had a problem for several years where the seatbelt warning beep would sound well after we’d started the car, later it was doing it more randomly and appeared to stop at some point. Belts always worked but other than an extended warranty on the driver’s door latch as part of a service campaign we never figured out what could have caused it. More annoying than a problem but Mom did see a newer Protoge in the service drive that was beeping constantly so it wasn’t just her car.
There was a rash of counterfeit Nichicon caps that made their way into all kinds of electronics. I just repaired an older Dell computer that is notorious for these caps failing and they are supposedly the extra long life Nichicon types we also use in a lot of our electronics. They are the only caps to have failed on that board (24 out of ~60 electrolytic caps) and if you look closely, there are differences in the shrink wrap and construction of the caps from legitimate Nichicon caps. Not easy to spot.