Went off to fix some nagging problems on vehicles involving electrons, the most pesky portion of the atom.
First is the '96 four-cylinder. It had a series of electrical problems that I figured had to be linked, so the question was, where to start? The car had an intermittent tach inop problem, an inop temp gauge, and a recently fixed AC that would work fine if jumpered but would not work through the car's controls.
I started on the tach, and fixed it by replacing the instrument cluster. I was hoping that would fix the gauge problem, but it didn't, dammit oh well.
Troubleshoot on the gauge led to the CTS. The CTS' on the post-'94's are a dual thermistor unit that has one thermistor drive the temp gauge and the other sends engine temp info to the ECU. Testing the CTS showed that it was opened up on both thermistors. Replaced CTS, and the gauge worked, and so did the AC. Apparently the internal ECU programming does not allow AC operation unless CTS is working correctly. So far, so good.
Off to the '94 3-cylinder. It had a temp gauge and AC problem as well. It also has radically different circuitry than the later models. For starters, it has no less than four thermistors/thermal switches driving various things underhood.
The temp gauge had the usual thermistor to ground circuitry, where varying resistance from the thermistor makes the gauge needle move. My gauge was stuck on cold, and that points to an open (one with no electrical continuity through it) thermistor. If the gauge is stuck on hot, that points to a closed thermistor. (one with no resistance through it). Testing the gauge thermistor, it was opened. $6 for a new thermistor later, it worked fine.
The AC has its own thermistor circuit in it for engine protection. If the engine is too hot, the AC is supposed to shut down automatically to keep from frying the engine. It does this through a thermistor that is a normally closed switch. If the AC controller sees continuity through this circuit the AC can run, and if it doesn't, the AC shuts down. The thermistor switch opens up at a some suitably high temperature. Testing this thermistor showed it stuck open, and so the AC controller would not send out an operate signal to the AC compressor.
Looking for this part was frustrating and I only could find it from Generous Motors, who wanted something stupid like $90 for it. I just fabbed up a short piece of wire with a spade connector on one end and an eye connector on the other. Run the eye connector under a bolt on the engine and hook up the spade connector to the connector on the harness and the AC controller will always see the cold signal. I won't have any engine overheat protection, but for $90 I'm willing to run that risk.
The other two thermistors on the vehicle are a thermal switch to turn the cooling fans on, and the engine ECU CTS. I figured that seeing as 50% of the thermistors on the vehicle were bad it made sense to replace them--they were cheap enough, less than $25 for both.
Latest Geo acquisition is this '98 3-cylinder I bought earlier this year. Previous owner diagnosed a busted transmission and wanted to sell it so I bought it. Fixed the 'broken transmission' with a new drive axle. Vehicle had signs of a spotty previous maintenance history, and so following my usual perfectionistic tendencies, I went ahead and went to stick in a new timing belt. While pulling off the front sprocket to change the front oil seal (I did say something about perfectionistic) I got well and truly bit in the ass when the sprocket drive key fell out. Looking at the keyway in the crank, I saw that it wasn't going to ever go back in, as the keyway was hammered out all to hell.
Discussed the matter with my machinist, Dennis Murray at Precision Engine Service here in Austin. Dennis has seen lots of these motors, on account of how a local (and very good) auto parts chain ran Geo's as delivery vehicles for over 20 years. (This chain since sold out to the large corporate conglomerate that aint no good. Sic transit and all that.) They had up to 50 of the things at one time, and a full-time mechanic to keep them running. Dennis said that he'd seen this problem several times before and that it seemed to be a common enough problem on a 3-cylinder Geo motor that had a lot of miles on it that ran the AC a lot. Here in Austin, most people run the AC 10 months or more of the year and I'm sure all delivery drivers do the same. Dennis recommended a crank replacement instead of a weld/remachine, but did point out how the mechanic at the auto parts chain used to fix them fairly successfully with J-B Weld. Secret was to clean things very thoroughly, and let the repair set up for at least three days before you turned the engine on. They'd gotten 50K out of some of the repairs like that.
I went off to the junkyard and found a good low mile engine from a car without AC and got its crank. Rebuilt the engine--with 182K on it it needed bearings, thrust bearings (there was a good 25 thousands of endplay out of that crank on account of the keyway hammering), valves and guides, and a set of rings. Engine should be good for another 200K, no problem.
Did a bunch of other work to the vehicle too, dammit, including a respray. Good running good looking car now.
But the thing didn't seem to be getting the MPG it ought to. When I went off to get the state inspection, the vehicle flunked on account of no data. Seems as Texas got rid of the sniffer as part of the emissions testing, and goes by the data stored by the OBD-II computers. (Not sure about what they do about the OBD-I cars out there--just look for a CEL on the dash and kick the thing out the door if it doesn't have one, I guess.)
Test station told me that this was a common enough problem if a vehicle had been worked on a bunch. Don't know if they are right on this--I got my doubts, and the folks running test stations aren't wrenches anyway. They told me to go drive it some more and generate some data for them to read.
But I had some doubts about this. It didn't sound right. So when I went home I got out the multimeter and tested the CTS. 'Sho 'nuff, the CTS was opened up, just on the computer thermistor side of it. Got and installed a new CTS and everything worked fine. Vehicle passed inspection with less than 5 miles on the new CTS. I think the car now gets 50 mpg on the highway, if I keep my foot out of it.
So there are two discoveries here. First is the crankshaft problem. Something people need to know about and be prepared for. Y'all up in Canada don't drive vehicles enough with the AC on to develop this problem, but I'd bet that most people south of the Mason-Dixon line do and will.
Second is the thermistor issues. I don't understand why neither the '96 nor the '98 did not throw a trouble code with a bad CTS. They should have, and I suspect some defective GM/Suzuki software engineering here. Another is that it looks like, at least in hot-climate places like Texas, that the thermistors in these cars are getting old enough to be failing in large numbers/percentages as is usually the case with electronic component failure of old electronic components. We all know the shape of that old electronic component failure rate over time graph.
I'm inclined to think that as a matter of troubleshoot routine and maybe as a part of routine maintenance it would make sense to check the CTS right off the bat. If you have a vehicle whose MPG's have taken a dive, CTS failure is a good probability, and they are easy enough to check, once you get the damned connector off for the first time. I'm also thinking that on your own personal vehicle it makes sense nowadays to just go ahead and replace the CTS as a matter of routine, now, if you live in a hot part of the country. I suspect my experience with the percentage of failed thermistors on my vehicles shows that they are starting to crater, and replacing one before it fails might keep you from melting an engine if you get stuck in traffic on a hot day with a dead CTS. It likely is better for all you up north, where cars don't run that hot underhood like they do down here.