My mechanic, who's an automotive electric system genius couldn't resolve why this would happen- Temp gauge is in the middle of it's range while running normally. When I turn on the parking lights, or the first click anyway, the temp hops up dramatically then comes to rest somewhat higher than before. Turn full lights on and the temp needle is just below going into the danger zone. The gauge returns to it's normal place when the lights are turned off.

Ideas?

 

A bad ground somewhere. It sounds like the temp gauge is sinking current from another circuit due to a bad or non-existant connection in a circuit related to the headlamp circuit.

I'd start with the ground connections related to the instrument pod and connections around the flex circuit board ... it probably wouldn't hurt to replace the flex circuit board because strange things can happen when it gets old.

This is a strange problem. If you find the cure, let us know! Good luck.

 

Does it behave this way when the turn signals are flashing?

The reason I ask, is that if the behavior is consistent between the headlights and flashers, then I would highly suspect an earth ground issue with the cluster. However, it may help to clean all ground connections..... Especially the braided cable between the engine and chassis.

I don't think the headlights will have an affect on engine temperature.
If you are highly concerned, use an infrared thermometer to verify temps before/after headlights.


I base my ground issue response on the fact that the park lights and headlights are drawing large amounts of currents from the chassis. If there is a high resistance path between the engine and chassis, then it could cause the ground reference point of the thermo. sending unit to drift ... Essentially an offset being induced that appears like a temp rise.

Hope this helps.

Scott


-Edit - Rich beat me to it

 

I agree - most likely a ground problem. If it's not a ground - 2nd check would be the "printed circuit" on the back of the instruments. A small short could be feeding juice from the headlamp (or other) indicator light to the temp gauge circuit.

If you haven't done so in a while - pull the pod - inspect the printed circuit, clean all the edge connectors, etc...

 

I have same problem as Joneser does, except temp needle moves up only 1/8 inch or so, and the fuel gauge also moves same distance. I've replaced the instrument circut, and cleaned all grounds. Not to sure where to go next. maybe replace the engine to frame ground strap.

 

From the wiring diagrams it looks as though the temp gauge and some of the light circuits run right next to each other where they connect to the central warning unit. Might be worth poking around in there looking for corrosion or other issues. I assume you've looked the fuse panel over front & back.

 

Fuse panel is o.k. Central warning unit is worth a look. Thanks for the tip.

 

Yes, we're pretty sure it's a bad ground situation and not worried about actual engine temp. Besides, I just went through the cooling system. New expansion tank, thermostat.

Turn signals do not cause movement of the gauge. It's too bad I didn't note this earlier, as I had the pod out right after I bought it to replace the ingition switch. We did inspect and clean what looked like potential trouble issues, but before then I hadn't driven the car at night.

Thanks for the feedback. I'll update as this progresses.

sparks- Willamette Valley? whereabouts? I'm in Albany.

 

Joneser - I'm in Salem. Are there many 928's in Albany? I've counted 9 in Salem.

 

This is how it went in my case: Voltage drop test revealed that it's too high at my firewall ground: 0.23V when ignition and light is on. At engine block, it's 0.06V the same time. As the temp gauge is grounded to chassis/firewall while it's sensor is grounded to the engine block, this ground difference must be the culprit of the temp gauge showing hotter temperature than what it is. No matter how much I cleaned the original firewall ground and how nice it looked like, the voltage drop remained high there. So I ran additional thick, 2AWG wire to the firewall from the negative post of the battery. Now the voltage drop at ground point is the same at the engine block, firewall, chassis and the frame rail next to the headlights as well even when the lights are on, 0.06V and my temp gauge stays steady turning the lights off/on. So this additional wire between firewall and battery seems to have solved my temp gauge issue..
I believe it has something to do with the instument cluster light rather than the headlights themselves as the cluster light probably share the same ground as the temp gauge which usually cause some trouble as far as I know..

 

My '86.5 has spiking temp guage also Reach behind cluster on left side. You will feel a plug with wires Push (pull actually) plug onto cluster. On mine the secureing tab is broken so it will not seat correctly at times. I think the outermost wire is the ground and would explain the spike. Also if spiked many times a firm tap on the outside will correct problem for a bit Hope this helps some

 

delamination of instrument panel flexible circuit board? I think there are about 6 ground "ribbons" and when the connection points start to turn to crap i would suspect that a positive (lights) on the female plug might be able to send a positive into the ground for the gauge....maybe? or, the ribbon with the actual signal power from the temp sensor is making contact and getting a full 12v when the lights are on? or the wiring into the female plugs has some missing insulation? too much power in the circuit from the sensor or bad ground are really the only reasons the gauge would peg, but the bad ground would be a consistent problem....since it happens when lights are on you have to be getting power from the light circuit to the circuit to the signal circuit for the gauge. But, if if the gauge reads higher with headlights on vs parking lights, and normal with no lights....that might blow up the circuit board theory.....i have to think about this one. (this is all a guess....but i ended up hardwiring my cluster because my flexible circuit board was absolutely trashed....i had all kinds of entertaining things happening with the dash lights and gauges...replacing the flexible circuit board with hard wire had changed my life completely. i am truly happy now)

 

Yes, in my case, the gauge reads hotter when lights are on, no matter if it's the head or the parking one so maybe it's the dashlight, but normal with no lights. The fans also cause the temp gauge to jump up when they turn on so it must have something to do with the electrical load in my case. But cleaning the grounds and running additional wires to the firewall seem to have solved my problem. I haven't tested it yet with the engine/alternator running as I also have other stuff to fix first but seems better and voltage drop tests on the main ground points also look better (with ignition and lights on).

If the lights/fans use the same ground return as the temperature gauge, it can explain the behaviour I believe. More current in this ground (because the lights/fans are are on) means more of a voltage drop in the ground. A higher voltage drop look to the temp gauge like a higher temperature. If the gauge would be grounded to the engine block as its sensor, or if the lights/fans couldn't use the same ground as the temp gauge, then I guess the gauge would be fine regardless of electrical loads..

Replacing the flexible circuit board is nice move

 

With electrical problems in the pod WE NEED TO KNOW THE MODEL YEAR because several areas in there went through substantial changes! i had the same problem,and it was a bad ground connection on the pod connectors. In early cars the connectors in the 3 hoods are quite soft, and I had bent some just by probing them with DVM leads. Turning on the side lights ticked up the temp gauge because the pod lights on indicator was gounding out through the temp gauge. On advice from here I re-tensioned the connectors with a dental pick behind the connector blade and all was normal again.
jp 83 Euro S AT 57k

 

In my case, it turned out that my dash lights caused the interference spiking up the temp and fuel gauge. If I took out the dash light fuse, the temp gauge dropped back even with the headlights on. I ran additional wires with ring terminals from the bolts on the back of the vdo instrument circled below to a ground point nearby and the problem went away. I guess this improved the ground of the temp and fuel gauges. These bolts that hold the gauges passed continuity to ground anyway..




​​