Final stretch on the 1983 944 n/a restoration with 77,000 miles. Car was idle for >10 years.


Replaced belts rollers, water pump (used old thermostat-ugh) coolant. Car runs and drives great but – Dash temperature gauge reads high.

On line graph of our gauge and reference points.


***f is Fahrenheit scale


Ambient temp around here, CT in the mid 40s to low 50s. When the car is moving temp is just shy of the three-quarter mark which I understand is 206°. Sitting it climbs just passed three-quarter mark, where the fans kick in. After a hard run it can climb almost to the overheat zone (221°) but not in it. Fans kicks in at high speed just passed 3/4 and after they run for about a minute or two temp moves back into its ¾ mark operating zone.

If the Fans are running the Left Fan will run on low speed if the car is turned off. .Fans operate at high speed when AC is on. I’ve taken pictures, infrared readings at the top bleeding elbow 194°, water pump inlet 165°, and on the single prong temperature sensor on the block173°. But Gauge indicates I am running206° to 221°. Not sure about calibration as my second infrared had lower temps. Bleeding Elbow Hot Out:

Water Pump Inlet-Hot in:

Temp sensor at block:



So my questions are: Is the car actually overheating? Is it a dash gauge problem(ground)? Temperature sensor problem? Thermostat? or even fan switch? What is the best way to approach this? Throw all those parts at it at once or do it systematically? Use lower thermostat and fan switch? About to put in a parts order and need to plan how I can attack this. Honestly based on past experience, touching the hoses etc. it doesn’t seem that hot in there.

Read everything I could find. Thanks in advance.

H

 

one thing to bear in mind is that the gauge sender is at cylinder 1 just, just after the water pump.
so it's reading the temp going IN to the engine...can only imagine the water coming out is scalding hot if its going in at 206.

try driving around with the fans locked on (turn the AC **** on, or short the fan switch wires) and see how the gauge responds with full-time airflow.

 

Sorry confusing. The blue numbers are Fahrenheit equivalent on the gauge. My temp at the pump is 165 and 173 at the sensor. I revised the drawing.

 

You might want to read this article: https://newhillgarage.com/2018/02/04...tem-explained/

As discussed in the article, the outlet and thus hottest water is at the top bleeding elbow. Remember, the system is pressurized so the boiling point is above 212 deg. More like 230 deg if you have a 7 psi cap on your overflow tank. So if you are reading 194 at the elbow, that's not a scary number at all. That said, the gauge on my track car rarely goes above the midpoint, except right after I shut off the engine after a session.

So I don't think you have an actual problem, except maybe with the gauge or the sensor. You can probably find the resistance of the sensor at 100C online and check it in a pot of boiling water on the stove. There may be high resistance in the wiring from the sensor to the gauge.

BTW, thermostats are not the problem or solution to overheating problems, unless they are truly stuck closed. Fans are for cooling at slow speeds.

Not part of your original question but I would put in a new Temp2 sensor (the one that sends a signal to the DME) as a matter of course.

 

Thank you. I read that article and found it very informative. So get both sensors? Hold off on thermostat and fan switch?

 

I would get the Temp2 simply because if they get wonky you can have some strange drivability problems. If you're comfortable with the fact that your gauge reads high, I might be tempted to leave it as is. A car that sat idle for that many years may have corrosion in the wiring in various places that you can deal with as you go along. You might want to start reading the forums about grounds, etc and start a preventive maintenance campaign to go in and clean up the various connections. The guys on the 928 forum are always fighting these issues. We recommend Deoxit. For starters, try pulling and spraying the 9 pin plug on your fire wall. A lot of important stuff goes through that connection (including your temp sensor wire) and it could be corroded.

 

My gauge was reading high on my 1988 924s. I ran an extra ground from the firewall to the engine and that fixed it. It may be just a ground problem.

 

I was under the impression a bad ground would have a sensor/gauge run low?

 

If the sensor has a positive temperature coefficient (resistance increases with temperature) then you would be correct. I did a quick internet search and could not determine if the 944 sensor is PTC or NTC. All the same, we all know that the ground paths can be just as important as the primary circuit. There are several grounds at the firewall just behind the rear of the cylinder head that need to be in good shape. But I agree with your thinking, the sensor reading high would be indicative of resistance in the primary circuit from the sensor to the gauge. A quick check is to ground the lead at the sensor and see which way the gauge goes.

 

I always wondered why the temp sensors in the WSM are called NTC I, NTC II, etc.

Just for kicks, bleed the cooling system again--for a long time. I had slightly higher temps for a few years after a coolant flush. The next time I changed coolant, I bled for extra extra extra long. Temps went back to normal

 

Dan, I never thought of it that way either. Could be. Or NTC stands for something totally German!

 

From Google translate: negativer Temperaturkoeffizient

Kinda close..?

 

On my 88S, cleaning the two grounds at the back of the engine dropped the temperature reading by ~ 1/2 of a division. If your gauge fluctuates when heavy electrical loads kick on and off (headlights, fans, etc.), that's a sign that you have resistance in the grounds interacting with the loads' ground currents producing voltage errors.

Temp sensors are negative temperature coefficient (NTC) thermistors - resistance goes down with increasing temperature. But this is not enough information to know which way the temperature reading would go since the thermistor must be in a voltage divider circuit to produce a voltage and the voltage coefficient of the circuit will depend on which side of the divider is the thermistor and which is the fixed resistor.

 

Good stuff Dan, Gtroth!!!

 

I have bled it two more times on an incline. No air. Temperatures last night (50 degrees ambient), building up at idle - 195 at the bleeding elbow 30-40 degrees going into the pump. Cleaned the main ground on the fire wall and the two behind the headlights. How many more are there. Gauge does not fluctuate with electric usage.