[andrewc24301]

Hi,

I have an 89 chevy celebrity with 2.8L engine.

I'm suspicious that the temp gauge is not accurate. These are my observations:

The temp gauge has a scale of 100-260. Mid point is 220. I've owned a celebrity in the past and it seemed to run around 220.

The car will go into the "red" area, of the gauge, it never pegs. However the car never shows any symptoms of running hot. Because they are affordable and easy to change, I popped in a new thermostat. It did not fix it.

I wrapped a digital temp probe around the thermostat housing an took a reading while the car was fully warmed up. I was receiving surface tempratures of 180-220, where as the cooling fan would cycle on while my probe was reading around 220.

Granted, this is not a very accurate way to measure, but it was the only means I had available. While I was taking this reading, the gauge in the car was borderline red. There is no number to define this on the gauge, but I estimate, the gauge was reading maybe 240-250? Somewhere in that neighborhood.

I replaced the temprature sender that was in the engine block. The first one I got was defective, the second one gave me the same readings, so I'm certain it's not the sender.

Now- here is my question:

It has come to my attention that there are to temprerature sending units in this car, one is plugged into the thermostat housing, and the other it plugged into the engine block. Ive reached the conclusion that the fan cycle relay must work off the one plugged into the thermostat housing, because it of the tempratures it operates in accordance with the gauge.

In my haynes book, wiring diagram, it also shows to ECT senders. One is connected to the computer, as well as the dash board "TEMP" light. And the other connects to the gauge. There is also a wiring going from that one to the computer it would seem.

I'm wondering does the sender in the t-stat housing control the fan switch for sure? And does this sender also control the computer functions, i.e. emmissions, when to run rich and lean mixtures etc?

In other words, what I want to know is - is the only functino of the sender located in the engine block just for the gauge? If this is the case, then could I just install a generic temprature gauge (the type that reads on a capulary tube)? Or would this throw the computer off. If so, how do I go about calibrating the on board temprature gauge for accuracy?

[andrewc24301]

I received no answer in this so I thought I'd update with what I have found this month.

After installing a generic gauge, I have concluded the original gauge is accurate.

I have determined that the sensor on the engine block reads to the gauge, and the sensor by the t-stat housing goes to the computer, which controls the fan.

What's interesting is that the fan comes on at 230 degrees, yet the gauges clearly go up towards 260. At these tempratures, the fan should cut on, yet it doesn't, I have concluded that the water temp at the tstat housing is not a 230 degrees.

So this is a brand new t-stat, couldn't be bad right? Must be air in the system. So I drilled a little hole in the t-stat to bleed the air out. No luck. Finally I just removed the t-stat, as now I'm beginning to suspect the water pump.

Well, when I removed the t-stat and drove it up drapers mountain, the temp never got above 150.

So I have concluded that the t-stat must have been bad out of the box. I have yet to replace it, thus far this month I've been driving without one, and the temprature has yet to go over 150 while driving. If it sets, it will go to 230, where the fan cuts in and it cools down to 220.

Since I drilled a hole in the t-stat I can not return it, however it was the cheaper one, I plan to purchase the "superstat" when I get paid and give that a try.

I will post my results.