drscottsmith

Mrmerlin

check the 14 pin connector a good chance the wire insulation is worn and the wire is shorting out ,
google Porsche 928 14 pin connector wire diagram

jpitman2

Why make it difficult by hiding your model-year? The electrics do vary from simple to complicated to digital-dash.
jp 83 Euro S AT 57k

drscottsmith

jpitman2

Measure the sensor - typically they are 3k+ ohm cold, dropping to ~2-300 ohm hot. If its ok, sounds like a wiring short.
jp 83 Euro S AT 57k

drscottsmith

All -

Resurrecting an old thread here with the same issues. Here is what I know now:

1) Coolant gauge immediately pegs when ignition is switched on
2) New coolant temp sensor installed
3) New front engine harness, 14 pin, etc. installed

Sensor measurements:
1) Body of sensor has continuity to ground (confirms ground path established)
2) Neither spade connector on sensor has continuity to ground (confirms sensor is not internally shorted)
3) Gauge side of sensor measures about 3xxx ohms when cold

Gauge behaviors:
1) Disconnecting gauge wire at sender results in needle dropping to dead
2) Disconnecting Plug 'G' on CEP results in needle dropping to dead
3) Connected properly gauge immediately pegs to top when Ignition is turned on

From what I understand, the absence of a signal from the gauge causes the needle to peg...so, here are my deductions:
1) The path to the gauge from the CE Panel must be ok since disconnecting at the panel causes the gauge to go dead. If there was a problem between the two, disconnecting at the panel would have no effect.
2) The path from the sensor to the CE Panel must be ok since disconnecting at the sensor causes the gauge to go dead. Again, if there was a problem between these points, then the disconnection would have no effect.
3) The gauge does not go through the Warning Box (only the light does), so that should eliminate any concerns there.

Any thoughts? Prior to this problem (and with the old harness and sensor in place) I went through a time when the gauge would act flaky - bouncing around and rising when it should not have. Finally it just pegs all the way hot.

Majestic Moose

That is very odd and as far as I can think you have ruled everything out. It could be that the gauge itself is shorted/defective.

drscottsmith

Moose that is where I am too - although that seems odd. I should add that I also have replaced the old foil with Roger's PC board kit, but the gauge behavior did not change after I did that either.

I would not have expected a gauge itself to go bad, but I certainly can't assume anything.

Does anyone have any idea how to test the gauge itself? I hate to pull the pod (again).

Jason89s4

Scott,
You've tracked down alot, but I don't agree with some of your assumptions. Let me see if I can explain why.

The absence of a signal to the gauge does not cause it to peg.
Your gauge operates on basic (+) and (-) voltage, but with the sensor providing just the right amount of resistance to ground to move the needle where it needs to be.
The provided (+) voltage is shared with other 12v(+) around it on the back of your instrument panel, within the printed circuit (or the whatever you have installed), including warning lights. If you are not having other gauge or warning light issues on the "left" side, then the (+) is probably okay.
The (-) side of the gauge goes from the back of the gauge, through the printed circuit, through "pin" 11 on the plug at the back of the cluster, to G4, to the 14-pin, and then to ground--but not before it goes through your sensor, which provides just the right amount of resistance to ground for the gauge to display the results. If you pull the wire off of the sensor and connect it to ground (thus, no resistance from ground all the way to the gauge) it should peg the gauge hot/high. So a gauge provided with the "least resisted" voltage indicates "hot". The gauge with "most resisted" voltage per the sensor indicates "cold". As jpitman2 said "3k+ ohm cold, dropping to ~2-300 ohm hot."

Your "deduction" #1 at end of post #6, while logical, may not be correct since unplugging "G" does more than disconnect the one wire from the temp sensor, it is disconnecting other items including (+) voltage that may be relied upon by the gauge to power up. Best thing is to measure continuity between the plug on the back of the instrument cluster at pin 11 and the plug at the sensor. That would test entire path from the cluster to the sensor. (Note that the plug on the back of the cluster is actually 14 pin, but only 2-13 are used. So, pin 11 will be the third from the end.)

All indications of a pegged needle when you connect the plug to the sensor and then dead gauge when removed from the sensor are that the sensor is grounded out. (I know you said you tested that, so that is the head-scratcher.)

Beyond a bad sensor, my guess would be the circuitry at the cluster plug or between the plug and the gauge. (That is where I finally tracked a similar issue with the temp gauge on my '83--more specifically the foil at the plug #11 was peeling away just enough to intermittently ground out with its "neighbor" foil.)

Also, you finally mentioned "Prior to this problem (and with the old harness and sensor in place) I went through a time when the gauge would act flaky - bouncing around and rising when it should not have." That behavior screams that a wire/connection/circuit is shorting out occasionally, and then finally grounded out. Are you sure the new wiring harness and instrument circuits are correctly installed.
Try something: switch the two male plugs at the sensor and see what happens.

I really doubt it is the gauge itself, because it is reacting to instructions...just not the instructions you want it to receive. You would need to access the cluster again to test the continuity of the circuits, and apply voltage directly to the gauge to eliminate that as an issue.
Let us know your findings.
Jason

drscottsmith

Thanks Jason and others! I will get to work and report back.

-scott

jpitman2

Have you looked into the pod connections? My temp gauge used to kick up a touch when I turned the lights on. Fitted a new sensor, no change. Finally I pulled the whole pod , did some cleaning of the contacts on the mylar, and in the connector hoods - worse errors. Took advice here - cleaning the connector blades inside the hoods is touchy - the blades are quite soft and bend easily. Solution - retension the blades with piece like a dental pick - fine point with a right angle at the end so you can get it behind the blade, and bend it towards the centre of the hood. All fixed - no more interference from one function to another reading, volts gauge is more stable (but not perfect). Along the way I found that the right hand hood(the 3 are labelled L, C, R) has multiple leads providing ground for the pod, and if the ground is poor, some gauges will misread as others use them as a ground path.
jp 83 Euro S AT 57k

Mrmerlin

what was the last thing you fixed before the gauge was working properly?

One of your comments was that you didnt want to pull the pod again ,
based on that comment you have probably pulled the pod off atleast once and done something on the installation of the pod to cause this issue.

Jason gave you a perfect system operation description and where to go to see the different signals and how to test them
NOTE do not leave the sensor wires connected the opposite way for more than a few seconds,
otherwise the internal coils of the gauge will overheat and melt.... BTDT.

Jason89s4

Stan,
drscottsmith mentioned in his first post that he changed the engine wiring harness. Then snuck in on post #8 that he changed the pod circuitry.
I'm going to circle back to MrMerlin tenet #1 "what was the last thing you worked on?" and I'm with you it looks like either the wiring harness or pod circuitry has some issues.
(Is why I suggested "Are you sure the new wiring harness and instrument circuits are correctly installed. Try something: switch the two male plugs at the sensor and see what happens.")
I've seen those two wires installed incorrectly. In '83 Porsche did make the error convenient--one wire is Blue w/ yellow, the other is Blue w/green. Can be pretty much indistinguishable after exposed 30+ years! Or colorblind.
Jason

drscottsmith

All -

will check everything. I used a pin out diagram when I built the harness (wire colors were definitely faded). But the issue with the gauge was there prior to the harness change - I was hoping the harness would fix it (due to resistance in corroded wires in the original harness).

will check though. When I get back I will have to check everything carefully.

Based on stan’s Advice...the problem was there with the old foil on the back of the cluster as well as with Rogers circuit board replacement. That was why I thought the sender and or harness could be at fault.

i am wondering if it may be an issue as mentioned in one of the pod connectors itself. Maybe a bent pin in the connector body.

will report back when I can get in there and have a look

thank you all

Mrmerlin

you may have to go as deep as taking the gauge apart to inspect it to see if the internals have melted, but this is a last resort.
The plastic coil holder inside the gauge will melt if the sender wires are swapped and plugged in for more than a minute with the engine running