Squirley Coolant temp

[Nitro Nick]

Unless there has been a failure, the stock wiring should work fine. After all, thousands of stock e30s use it without issue.

As others have suggested, you should test the sensor before simply replacing it. Measure the resistance at room temperature, in an ice bath, and in boiling water. Make sure to keep the connector area dry as the water might throw off your resistance reading. The readings you listed before were probably an order of magnitude off or so--as in the multimeter was auto-ranging. You need to note if it says Kohms(kiloohms) or anything other than just ohms.


solder is super brittle, your fix may have cracked and come apart. A high quality crimp with butt connectors is more reliable. Thicker or thinner gauge for this application won't matter too much. Even 22ga is fine. The factory size is roughly 25ga wire.

With the ecu disconnected, disconnect the coolant sensor and connect a jumper between the two female pins on the harness. A paperclip will probably work. Then, with a multimeter, test for continuity and resistance (ohms) between pin 45 and pin 2 at the ECU. Also make sure the value is roughly the same as pin 45 and pin 14, as well as pin 45 and pin 24. Have a friend wiggle the harness at the motor and c191 while reading the value across the pins to see if there is a poor connection.

pins 2/14/24 are all on the ground for the coolant temp sensor. Coolant temp signal in is pin 45.

Thanks jax. Ill have to try this test. I got a new sensor and did some of my own testing with the ice water bath, ambient and warmed up temps. I think my numbers were relative to the drop down selection that nando is referring to. When I tested the one currently in the car before the new one was installed the ambient temp ohms were off the chart, which im thinking may have been a bad sensor. When I put the new one in and re-calibrated it seemed to be working well BUT. when the car was warm the same thing occurred. The MS gauge would read 100 Degrees F at fully warmed up temp (roughly 176oF) and during driving it would go back and forth from 100 to 176 on the MS table gauge(during these trips the cluster gauge read perfect warmed up temp). Which leads me to think there is some sort of issue with the wiring. Whether that be a ground or the signal wire. So I am going to perform the test you suggested and see how it goes. Will that test identify if its the signal wire or a grounding issue? Or will that just tell me in general if its a wiring issue...

Thanks again...i'm almost there with this lol

[Jaxx_]

The gauge cluster uses a separate circuit and sensor to show temperature.

Regarding your follow up question: the coolant temp sensor is kinda an independent circuit from the rest of the grounding plane. In essence, the ground for the sensor is at the ECU(pins 2,14,24). The signal wire goes from the ECU to one side (doesn't matter which) of the CLT connector and then the ECU looks at the resistance between pin 45 and the ground pins. If a connection is bad and/or disconnects, the resistance will rise. As you might have seen from your tests in an ice bath and boiling water, the sensor reads a higher resistance (more ohms) at lower temperature. If the connection has failed or is near failure, you would expect temperature to jump to lower values than actual (176F to 100F, for example).

If you discover there is a connection issue with the tests, you can then isolate one side of the connector to test ground, and the other side to test the signal side. You need to identify the signal side of the connector and the ground side of the connector, then test them.

Do this:
disconnect the ECU.
disconnect the coolant temp sensor.
Find continuity between the ground pins (2/14/24) on the ECU and the ground wire on the connector to the coolant temp sensor. It will be one of the connector's pin receptacles or the other.
Once finding continuity for the ground pins to the CLT connector, you can then assume that the other pin receptacle on the CLT connector is the signal wire.
Find continuity between the signal pin (45) on the ECU and the signal pin receptacle on the connector.

Once you have found continuity at both, you want to confirm continuity for each circuit as you wiggle the harness across the engine bay, especially where you've made your repairs. ANY breaks in continuity or changes in the resistance not caused by your multimeter leads losing connection means that there is a break in the wiring harness for that circuit, and needs to be repaired. Remember that just because you find ONE break in continuity does not mean that fixing that will solve the issue. The other side of the circuit must also be confirmed.

I hope that makes sense. If you can't find any issues with the car cold, plug everything back in and run the motor until the issue appears and test again. If you have alligator clips or an extra pair of hands, it helps a lot. Someone to hold the DMM while you wiggle the harness or vise-versa. If you DMM has a mode of audible buzz/beep for continuity, it will help too.