The last time I thought my temp gauge wasn't right I re-seated the connector on the block sender. The temp wasn't coming up according to the gauge, it was floating down in the cool range. After my sender repair, The gauge was running steady and stayed on "N" on my daily commute once the engine was warmed up and the t-stat was operating.

I have been very happy lately about the readings from my gauge. Then an interesting emergency happened last Saturday afternoon on an 85 degree day. This little story may provide some clues in the wiring versus air bubble debate. I am neutral on the subject, and am not trying to state that I know why anyones' temp gauge might be erratic.

I recently had a major coolant leak from the radiator, the plastic POC draincock stem busted in half and left me stranded. I was parked at a convenience store and discovered coolant flowing out of the radiator when I returned to my truck. I made a decision to try and make it to the closest auto parts store, knowing that I was risking some major damage to my new engine( only 18,000 miles on her). I limped to the local auto parts store, (I just made it, almost overheated the engine) and replaced the draincock assembly and topped the cooling system off with 10 quarts of 50/50.
I ran the engine in the auto parts parking lot till it was warm, topped off the reservoir and headed home. During my 23 mile drive I stopped once to check the reservoir and inspect for leaks. I tried purging some air once by squeezing the upper hose and made it home OK. The next day I added a few pints to the reservoir and the system has been OK for one week now.

When the leak occurred, I was missing most of the coolant and had no draincock, the gauge acted normally, that is, as you would expect it to operate when 10 quarts of coolant are missing. While on the 2 mile run to the auto parts store, I assume there was alot of air in the system, the
gauge started moving up fast and reached the top scale very quickly, and stayed there, it was on the"L" or a little higher. The temp gauge told me what I was almost 99 percent sure of, that I was in danger of overheating! I was shutting the engine off and trying to glide in neutral gear when I could, because of the heat. I bet you could fry a 4.9L in less than 5 miles, without any coolant, it was a scary ride. I was prepared to stop every quarter mile if I had to and let the engine cool, but I made the 2 miles in one manuever.

After the coolant was replaced, I drove home, and the gauge was steady at "N" all the way. It has been OK for a week now, and I haven't had to make any special effort to purge trapped air.

The sender in my truck worked just fine, or at least gave me readings I would expect from an engine full of coolant, and then the next minute leaking coolant like crazy. If the air bubble theory is correct, then why didn't my temp gauge start acting erratically while most of the coolant was missing?

 

Because you have a stock gauge, it doesn't show the fluctuations because it's not nearly as reactive to extremely quick flucuations in temperature, keep in mind I have both the autometer gauge (which DOES show you those flucuations) and the stock guage (which doesn't) both tapped into my motor. I may have confused you, and maybe at times you thought I wasn't specifically refering to an autometer gauge.

If you took that stock gauge setup and stuck it in a pot of hot water, then quickly took it out....do you think that needle would bounce to H and back to L? No, it wouldn't...it's not reactive enough. If you did that with my autometer, it would show you the temp as soon as you put that gauge sending unit in the hot water, and go back down when you took it out.

You can't even see the t-stat open on a stock gauge...but you can on the autometer because it shows you exactly what's happening and the stock gauge is not setup to show you exactly what the temp is at any given moment.

The stock gauge does not necessarily tell you what the temperature is at any given moment. It does lie to you

All manufacturers set it up this way (to some degree) because the average driver just wants to know that everything is ok...and thats all the gauge will tell you.

I appologize if I am coming off as a "know-it-all", but I am just trying to work through this like you guys are, and I feel that I've made a lot of progress and my theories seem to fit. If it wasn't so dang hot out here I'd go work on the thing.

 

If you're sure it's an air bubble in the cavity where the sending unit screws in why not just loosen (unscrew) the sending unit a little, start the engine, and purge the trapped air plus a little coolant. I suggest doing this while the engine is cold for the obvious reason. It would be best to retighten the sender with engine running to prevent the possibility of air getting back in (although not likely). If you want to check the gauge circuit itself remove the wire from the sending unit and you should have no gauge deflection. Short the wire to ground and you should get full deflection. If you want to go even further, go to a electronics store and get a 10 Ohm 1 Watt resistor to put in series with the sending unit wire and ground, this should cause the gauge pointer to move to the highest marking on the gauge face. 73 Ohms should make it read at the lowest marking (75 Ohms is the closest standard resistance value you can likely find but that's close enough). This is also true for oil pressure and fuel gauges. Erratic readings can be caused by instrument panel voltage regulator.

 

Well, it seems as though it's working.

I havn't touched the truck since about 3 weeks ago when I got a little air out of the radiator. I didn't really do anything in particular to get the air out, I was topping it off and some came out. I've driven it twice since then, and both times the temp on my autometer was working perfectly about 90-95% of the time. Previous to that, it was working about 40-50% of the time. Once or twice i've seen it drop, but everything is lookin' pretty good now.