92 4.3 temp sender for gauge error

One day I noticed the temp on the dash was showing totally cold, even tho the engine was warn and the heater blowing hot air! My 92 s10 blazer with the 4.3 Vortec CPI had an issue. I checked the sender for the gauge and when I pulled the connector off the whole end had rotted off the sender. I put in a new sender and now when the ecm (or whatever its called) times in the temp climbs up slowly till its pegged out and the check gauges light comes on the dash. It measures about 3K ohms ambient. So I went to a dealer thinking aftermarket was off some.... measured the dealer item and it was the same. I guess too low resistance measures hot all the time. That did not help. I put a 10K ohm variable resistor in place of the sender and was able to set the gauge reading wherever I wanted so it seems that the circuit is working. I wanted to install the variable resistor under the dash instead of under the hood but am unable to find the green wire on the ecm connector. Can't seem to locate a wiring diagram of the sender so unless someone has one handy I'll be mounting it under the hood!

hMMM, MORE info is better. But in this case it confuses me. I checked the value that resulted in a 180 degree indication on the dash (140 ohms) using my
variable resistor (POT for short). I checked the value of the sensor while the engine was slightly warm (1000 ohms) which indicated totally pegged out to HOT. Turning up the POT to 300 ohms lowered the dash reading. So how does increasing the POT from 140 to 300 ohms lower the reading but 1000 ohms on the sensor causes a High Temp pegged indication. I thought it would read COLD.

So I'm at a loss, maybe I'll take a heat lamp and thermometer to my local parts store and see how one acts on the bench...
(aprox. values)
Open circuit = COLD
140 ohms = 180 degree temp
300 ohms = 80 degree temp

But 1000 ohms connected to sensor = HOT

two connections on sensor - only one wire on connector. This is strange, If I come up with anything I will post, just FYI!

This is the customary resistance/temperature scale:

http://www.wwdsltd.com/files/CTS-MAT.gif

That seems to bear out my results with the POT. There must be some issue with the sending unit I installed. Somehow when in actual use it must have the wrong value.... I will get another one and try that! Thanks for the list!

A new sensor fixed it! Works like it should. Some issue with the other sensor under working conditions must have resulted in a much lower resistance than was measured with the ohm meter ...