Temperature gauge not working and cooling fans run all the time?

Upon looking at the wiring diagram, your model uses a pair of radiator cooling fans, side by side, that are controlled through three relays. The relays are shown in the first diagram. The second drawing shows the test steps provided by GM. See if you can determine by this chart if both fans are running on the low or high speeds. I'm trying to figure out if the Engine Computer is turning the fan relays on in response to a sensor reading, a switch setting, or on its own with no regard to the switches and sensors.

Without scanner data to look at, the only other test I can suggest is to monitor the CTS voltage again on the yellow wire as the engine warms up. Remember, this has to be done with the sensor plugged in, by back-probing next to the yellow wire.

The voltage should be somewhere near 4.0 volts when the engine is still cold. As it warms up, the voltage will drop, but a surprise will occur when the coolant reaches roughly 160 degrees. By that time the voltage will have dropped to around, ... Oh, ... Roughly 1.7 volts, or maybe 2.0 volts, then jump back up to something higher, ... Say 3.0 volts for this story. At that point the voltage will resume dropping as the coolant gets closer to 195 to around 212 degrees or more.

The CTS circuit is fed with a carefully-regulated 5.0 volts from inside the computer. All of that 5.0 volts has to be used up, or accounted for. Some of it is "dropped" across circuitry inside the computer. What is left is what you measure on the sensor's yellow wire as the signal voltage.

To achieve greater accuracy when coolant temperature is close to the normal range, different circuitry is switched in inside the computer. That reduces how much of that 5.0 volts is dropped in the computer, and increases the amount left over to be seen at the sensor. That's the point where the signal voltage pops back up a couple of volts, then resumes its gradual decline as temperature goes up.

If you find that odd, but normal operation, the CTS circuit is working properly. Then it becomes an issue of figuring out if the computer is responding to something else.

I just saw that too. Don't know why the first two drawings didn't show up.

If they're on slow speed, you can unplug Fan Control Relay # 3 and nothing should change. Current is flowing through first one fan motor, then the other fan motor, then to ground. With both motors in series, each one gets only 6 volts, so they run at half of normal speed.

To do that, Fan Control Relay # 1 has to be turned on, and for that to happen, pin # 6 at the PCM has to be "low", meaning the ground path is switched on. That energizes the coil inside that relay, and it clicks on.

Just so we don't overlook something stupid, you might want to switch the locations of the relays in case one is sticking on. Suggest switching relays 1 and 3. If you feel the relay click when you plug it into socket # 1, that's further proof the computer is turning it on.

I expected the CTS voltage to go to "0" when the ignition switch was turned off. I never checked, but the CTS voltage is not looked at by the computer once the ignition switch is turned off, so I can't say if what you found is normal. When the computer runs the fan with the ignition switch off is strictly a time thing, not a temperature thing, at least that is what I was always told. If you want to prove this to yourself, nothing you've observed should be different if you unplug the CTS after the ignition switch has been turned off.

Under normal operation, when the switched ignition 12 volts is turned off, the computer immediately turns off its 5.0-volt power supply for the sensors. Your system operates differently in that it stays "awake" in off mode to run the circuitry that runs the fan relays. That implies it could indeed keep the 5.0-volt supply up and running for a few minutes.

Have I confused you yet?

Voltage is supplied to the relays with the ignition switch off. By the mid to late '90s, we were calling that, going to "sleep" mode. That can take up to 20 minutes on some models.

It appears what is happening on your model is the computer does stay awake, and it still runs the sensor circuit during that time. The idea of running the fan is it cools the coolant in the radiator, and the temperature difference to the coolant in the engine causes a small amount of flow even though the water pump is not running. The other point of interest is the sending unit is some distance away from where the heat is generated within the engine. It can take some time for the higher temperatures to migrate over to the sensor. That will be seen as a temperature change, with corresponding signal voltage change, over a short period of time, with the engine off.

Another school of thought is when a hot engine is stopped, there's still a lot of heat in the engine's cylinder walls and head that continues to migrate into the coolant. With no air flowing through the radiator, the warming coolant in the engine expands and pushes some out to the reservoir, but the heat remains where it was generated. That might lead to a cylinder head warping, resulting in a leak. By running the fan a little, the coolant in the radiator contracts. The expanding coolant in the engine flows to the cooler area, thus drawing some of that heat away from the cylinder head. That high heat in the block can raise coolant temperature higher than is normally seen when running, due to the lack of a little flow. That occurs most when a hot engine is freshly stopped, but rather than try to get involved with complicated software to figure out when it's necessary to run the fan, they just build the system to run for one minute each time the ignition switch is turned off.

My assumption at this point is the CTS circuit is working properly, and just the stepper motor for the temperature gauge needs to be replaced. The best way to see the CTS operation is by watching the voltage readings on a scanner. The second-best way is with a voltmeter, like you did.

That's the time issue that I think is built in. It's not really related to outside temperature or coolant temperature.

Not all auto parts stores use scanners. They often can just read fault codes. You'll have to ask how deep they go into diagnosis. Gotta remember they're in the business of selling parts, not repair services.