Replace Engine Coolant Temp Sensor or ECM?

[Karms20]

On our '94 Bonneville SLE (3800 engine) after some engine warm-up, the check gauges light illuminates for approx. 5 to 10 seconds and I hear a chime, but no issue with the temp, oil, or voltage gauges. Also, the electric engine cooling fan(s) are "on" when the check gauges light is illuminated.

Using an OBD Diagnostics cable connected to the vehicle's 16-pin ALDL connector and my laptop, I monitored the engine coolant temp sensor. TunerPro RT shows the sensor temp increasing from a steady 154F to 296F in just 0.2 seconds. The temp was then erratic mostly above 250F for 5 to 10 seconds before dropping back down again in 0.2 seconds to the steady 154F. The engine cooling fan(s) were also "ON" in the data log when the temp was high.

I'm trying to determine if the sensor or ECM needs to be replaced. I'd rather not remove the difficult to get to sensor and test it if I can avoid it. I also know that this vehicle is supposedly notorious for ECM failures. I replaced the ECM 12-years ago (32K miles). Current mileage is only 81K.

Any help would be greatly appreciated. :mad1:

[sturgillbd]

Considering the CTS drops in resistance as the temp rises, I would check the wiring to the sensor before I replaced anything. Look for a pinched wire etc. The sensor could be shorting internally also. If the sensor is unplugged, the coolant temp usually shows -40 degrees on a datalog or scantool

[Karms20]

Sturgillbd, thank you very much for your response.

I didn't notice any issues with the wiring.

Which CTS supplier(s) would you, and anyone else, recommend from experience? Standard? Echlin? Welds? AC Delco? Borg Warner?

[Six_Shooter]

Could it be something as simple as low coolant?

I've seen very similar jumps in coolant temp just from low coolant, where a pocket of air (always hotter than the coolant) is then surrounding the sensor for a brief period.

You could check the voltage on the sensor signal wire as the engine warms up and if there is a sudden change that corresponds with the other symptoms, I would go ahead and replace the sensor. If there is no coinciding change in voltage than I would look elsewhere.

[bybyc5]

Sure sounds like a failing CTS.
The ECM is just monitoring the voltage drop in that circuit. Anything is possible though.

Anyway to put an CTS in another location, and just rewire to it? :idea: I'm not familiar with that 3.8L motor and it's cooling system layout. Maybe remove the temp sender for the gauge and put a new CTS in that hole for test purposes.

Just where is the coolant temp sensor mounted?

Unplug the connector if you can get to it and temporarily push in a radio shack resistor of what ever the value is for 155*F and see if it reads stable all the time. I like the steam Idea, save for the fast reaction time of .2 seconds to respond.

What part of the ECM is known to go bad in those ECM's?

[Karms20]

Could it be something as simple as low coolant?

I've seen very similar jumps in coolant temp just from low coolant, where a pocket of air (always hotter than the coolant) is then surrounding the sensor for a brief period.

You could check the voltage on the sensor signal wire as the engine warms up and if there is a sudden change that corresponds with the other symptoms, I would go ahead and replace the sensor. If there is no coinciding change in voltage than I would look elsewhere.

The coolant level was fine. The radiator was full. No air detected when bleeding the system.

This weekend I replaced the engine coolant temperature sensor with a new AC Delco part. Unfortunately, I noticed basically the same jumps in coolant temp when data logging.

Is it possible for this erratic behavior to be caused by an intermittent wiring issue? It seems to me it's now either wiring or the ECM. (Unless the new part is defective. I hope that's very unlikely.)

[Karms20]

Sure sounds like a failing CTS.
The ECM is just monitoring the voltage drop in that circuit. Anything is possible though.
Anyway to put an CTS in another location, and just rewire to it? :idea: I'm not familiar with that 3.8L motor and it's cooling system layout. Maybe remove the temp sender for the gauge and put a new CTS in that hole for test purposes.

Just where is the coolant temp sensor mounted?

Unplug the connector if you can get to it and temporarily push in a radio shack resistor of what ever the value is for 155*F and see if it reads stable all the time. I like the steam Idea, save for the fast reaction time of .2 seconds to respond.

What part of the ECM is known to go bad in those ECM's?

The CTS is mounted underneath the throttle body. Wasn't able to get to connector to push in a resistor before replacing with new AC Delco sensor. Blind removal & install.

That's a very good question about the ECM's. Could something like the 5 Volt reference voltage to the sensor be fluctuating intermittently?

Is there something I can look for in the data logs to pin down that the ECM needs replaced?

[bybyc5]

Well there is only 3 things it could be..The sensor, the wiring, or the PCM. You have ruled out the coolant sensor. The harness can be the next easiest ( well cheapest anyway ) thing to verify.
Odd that it jumps to a high temperature, indicating a short to ground somewhere in that circuit.
If you can individually isolate the circuit wires and test them from the PCM connector to the Coolant sensor connector for continuity, with no pinched wires intermittently shorting to ground somewhere, then it must be inside the PCM.

Sometimes it's easier to overlay temporary wires from connector to connector if the circuit continuity test does show a problem and the harness is long and buried or routed in such a way that exposing the problem is not easily done. It's hard to visually inspect every inch of the suspected wire for failures..shorts or open circuits.

If you can duplicate this event with any regularity, unplug the coolant sensor and run the car while monitoring that sensor circuit to see if the glitch is still there. If so you know the problem is the PCM. Only down side is the engine will run very rich with the coolant sensor unplugged. That monitored value in TunerPro should read -40F or so unplugged.

[Karms20]

Well there is only 3 things it could be..The sensor, the wiring, or the PCM. You have ruled out the coolant sensor. The harness can be the next easiest ( well cheapest anyway ) thing to verify.
Odd that it jumps to a high temperature, indicating a short to ground somewhere in that circuit.

The vehicle has the original spark plug wires (21-yrs old, 80K miles). Is it possible the spark plug wires now have deteriorated/leaking insulation allowing sparks to intermittently leak through the insulation to the engine coolant temperature sensor wiring creating an induced voltage?

This intermittent increased voltage would then be "seen" as an intermittent increased engine coolant temperature by the ECM.

[Six_Shooter]

While I don't believe that is what is happening, I would definitely replace those 21 year old spark plug wires.