Temp Gauge

[Radio-Active]

Well, hmmm...

Hang on a sec... I'm thinking that won't work as the sensor is a path of variable resistance to ground. There has to be a voltage source somewhere and I don't have the miata wiring diagrams on this computer in the shop. What you'll end up reading is a jumping voltage source from somewhere up the wiring from that variable ground point.

Trace it back in the diagrams. You may just have a bad connection on the instrument cluster or a bad wire somewhere else. When the car warms up can you slap the dash and get it to jump all around?

RJ

Did'nt they ever teach you not to second guess yourself in school?

The temperature sending unit is made up of two dis-similar metals that create their own voltage by way of, for lack of a better term controled corrosion. The heat applied increases the amount of voltage by the two metal speeding up the 'corrosion process. IOW Hook up the positive to the place on the sending unit and the negative on the ground(unless it is a two wire sending unit) it will probably peek out around 5 volts.

[cam76034]

It has now been sitting in the driveway idling for about half an hour. Gauge at 11:00, fan is coming on every so often. In other words, I can't get it to overheat under normal overheating conditions.

[trickyrix]

From a similar experience with my BMW, I'd say the problem is in the gauge cluster. The solder joints on the pins that plugged into the back of the gauge would get corroded and crack, causing some funkiness in the contact at that point. The gauge would whip back and forth, then settle down... or go dead... whatever it felt like. I could sometimes whack the dash and get it straightened out... or not. I just learned to live with it after replacing the whole cluster twice (junkyard bits, of course).

Sounds like you don't have an overheating problem... and maybe the previous owner never had an overheating problem. I've heard some pretty goofy diagnoses from those that don't know what they're talking about...

[cam76034]

Wish I had an extra gauge to test that theory. It's unnerving to try to drive with the needle pegged past H.

[trickyrix]

Wish I had an extra gauge to test that theory. It's unnerving to try to drive with the needle pegged past H.

Um... don't you have TWO extra CARS???

[cam76034]

Okay time to approach this from another angle since voltmeter test was inconclusive. Would one of those infrared thermometers aimed at the engine give an accurate temperature of what the coolant temp is? Or maybe a meat thermometer sitting on top of the manifold while driving?

[trickyrix]

Okay time to approach this from another angle since voltmeter test was inconclusive. Would one of those infrared thermometers aimed at the engine give an accurate temperature of what the coolant temp is? Or maybe a meat thermometer sitting on top of the manifold while driving?

I guess you could train an IR thermometer on the upper hose or thermostat housing and get a reading. You might not get an accurate temp reading (but you could probably extrapolate one). At any rate, if you had a cooling problem, I think you should be able to see any wild fluctuations in temp. If you didn't see any huge changes, you could probably rule out a cooling problem...

At least it makes sense to me...

[cam76034]

I guess you could train an IR thermometer on the upper hose or thermostat housing and get a reading. You might not get an accurate temp reading (but you could probably extrapolate one). At any rate, if you had a cooling problem, I think you should be able to see any wild fluctuations in temp. If you didn't see any huge changes, you could probably rule out a cooling problem...

At least it makes sense to me...

Problem solved!? At least I'm about 90% sure. I had taken the instrument cluster out and probed voltmeter into the lead for the temp gauge, but didn't really get anywhere that way. So I went to the 99 cent store and bought a 99 cent meat thermometer. Then I wedged it in between the head and exhaust manifold. Drove the 2.5 miles to the nearest highway and checked temp at 175 degrees. Then I got on it up and down 121 several times, but the needle just stayed at 11:00! It just wouldn't budge, and I was doggin' it. Temp still at 175.

I think it was the gauge, or connections thereof, and it looks like taking things apart and putting it back fixed it. Trickyrix was probably right, most likely never had an overheating problem. Attempts by previous owners showed they had tried all the usual right things, all new hoses, the temp sender unit had been fiddled with, and there's a crack in the plastic across the instrument panel where I'm guessing somebody was trying to get to the gauge, or was banging on it till it cracked. I'm not sure but it looks to be a fairly new radiator as well. Kind of like a puzzle where everything falls into place.

[trickyrix]

Trickyrix really knows his shit*, most likely never had an overheating problem. Attempts by previous owners showed they had tried all the usual right things, all new hoses, the temp sender unit had been fiddled with, and there's a crack in the plastic across the instrument panel where I'm guessing somebody was trying to get to the gauge, or was banging on it till it cracked. I'm not sure but it looks to be a fairly new radiator as well. Kind of like a puzzle where everything falls into place.

That means you've got $800 worth of new parts and labor already done for ya!

Hope that was the only problem... if so, you're a lucky dog and got a hell of a deal! BTW, those instrument hoods are about $100-120 - I'd JB Weld the sucker and paint it!





*Of course, I had to embellish my role in this...