Fuel gauge problem

Talk and Tech about turbocharged 924/944/968 cars
User avatar
Belgian951
Posts: 224
Joined: Tue Oct 03, 2023 3:59 am
Has thanked: 74 times
Been thanked: 33 times
Hi everyone,

I have the quite often talked about fuel gauge problem on my 1986 944 turbo. It's keeping me from driving the car since not knowing how much fuel you have kinda makes it annoying. I need to fix this issue for my peace of mind.

So I did the Clark's Test (found here: https://www.clarks-garage.com/shop-manual/elect-13.htm) and hooked up a potentiometer and set the resistances found in the late model table:

RESERVE: 63.2
1/4: 42.2
1/2: 21.2
3/4: 8.6
FULL (1/1): 2.8

My readings however did not align to the ohm values in this table (or the early table). It seems like I have to increase the ohms by quite a bit to get the desired fuel gauge levels.

RESERVE: 85 ohm, but reserve light didn't light up, needle def on empty position
1/4: 68.4 ohm
1/2: 46.7 ohm
3/4: forgot to test it seems
FULL: 17.5
OVER FULL (needle a noticable tad over the 4/4 position on the gauge): 2.8 ohm (which should be the test value for normal FULL needle position)

If I remove my test set up, put the car together again, and start it, the gauge shows 3/4 full. Actual gas level is unsure.
Any tips of where the issue lies?

My test set up:
WhatsApp Image 2025-02-08 at 23.25.29.jpeg
WhatsApp Image 2025-02-08 at 23.25.29.jpeg (330.34 KiB) Viewed 5572 times
1986 944 Turbo Garnet Rot Metallic

#1

User avatar
Tom
Site Admin
Posts: 8938
Joined: Fri Jun 25, 2021 2:04 pm
Location: Silicon Valley, CA
Has thanked: 934 times
Been thanked: 4005 times
Contact:
I'd take the gauge out of the cluster and check it directly. Some of those gauges have a trim pot on them, and some don't, so if you are lucky you can make it accurate just by adjusting the pot. Probably not, but worth a try.

To test the gauge, you need to insert something into where the connector pins normally plug in. I use molex pins and had to squeeze them a bit to fit snug in the gauge ports. Red jumper is 12 volts. Black is ground. White goes through a variable resistor to ground. Idea is to turn the resistor until it reads full, half, etc., then disconnect and see if the resistance is within spec for those readings. Cleaning the edge connector and football connector can lower the resistance seen at the gauge, but it sounds like you have the opposite problem. So if yours is off and the trim pot doesn't help, then something's wrong with the gauge. There is a resistor that runs over the top of the trim pot, and the solder joint can pop loose, causing roughly your issue. You can see it in the pics below on top of the trim pot. Other than that, you can also check to make sure the needle is oriented properly and wasn't just removed and replaced so that it points way too high. It should rest near empty and stop near full if you nudge it with your finger.


Image 11.jpeg
Image 11.jpeg (80.87 KiB) Viewed 5558 times

Image 14.jpeg
Image 14.jpeg (132.1 KiB) Viewed 5558 times
Image 12.jpeg
Image 12.jpeg (49.98 KiB) Viewed 5558 times
Image 13.jpeg
Image 13.jpeg (293.65 KiB) Viewed 5558 times

#2

User avatar
Belgian951
Posts: 224
Joined: Tue Oct 03, 2023 3:59 am
Has thanked: 74 times
Been thanked: 33 times
Thanks for the detailed explanation! So you think it is probably a gauge problem and not a fuel sender problem, correct?

I was thinking it could be a double problem. The gauge doesn't send the correct ohm (since fuel level isn't moving during use, if I recall correctly), but the gauge also doesn't react correctly to the delivered ohms (since I had to add more resistance during the test).

Good suggestion to do the test directly at the gauge. That'll eliminate a wiring problem and a sender problem at the same time.

To be clear, the trimpot is shown on the left side in your first photo?
1986 944 Turbo Garnet Rot Metallic

#3

User avatar
Tom
Site Admin
Posts: 8938
Joined: Fri Jun 25, 2021 2:04 pm
Location: Silicon Valley, CA
Has thanked: 934 times
Been thanked: 4005 times
Contact:
Belgian951 wrote: Sun Feb 09, 2025 1:02 pm Thanks for the detailed explanation! So you think it is probably a gauge problem and not a fuel sender problem, correct?

I was thinking it could be a double problem. The gauge doesn't send the correct ohm (since fuel level isn't moving during use, if I recall correctly), but the gauge also doesn't react correctly to the delivered ohms (since I had to add more resistance during the test).

Good suggestion to do the test directly at the gauge. That'll eliminate a wiring problem and a sender problem at the same time.

To be clear, the trimpot is shown on the left side in your first photo?
What symptoms are you seeing from the gauge? Reading too high? Too low? Not moving? Just FYI the gauge does not 'send' ohms to anything. It only reads the resistance (ohms) on its input terminal. Think of the sender in the tank like a dimmer switch. The more you dim the switch (i.e., the more ohms it makes), the lower the gauge reads. It sounds like your gauge reads higher than it should, for any given setting on the dimmer switch. Usually it's the other way around, where the harness and cluster connections create extra resistance and lower the reading on the gauge. That's why my hunch (just a hunch) is that the issue is in the gauge itself, and why I'd pull the gauge out of the cluster to (a) check the trim pot and make sure that resistor hasn't come loose, and (b) test it on its own as shown above.
fuel-gauge.jpg
fuel-gauge.jpg (519.74 KiB) Viewed 5505 times

#4

User avatar
Belgian951
Posts: 224
Joined: Tue Oct 03, 2023 3:59 am
Has thanked: 74 times
Been thanked: 33 times
Thanks again for your reply. I meant that the fuel float (sender) doesn't send the correct resistance to the gauge, and that the gauge also doesn't react correctly.

During driving, I'm seeing no movement of the gauge IIRC (I haven't driven much yet), but the low fuel indicator light did light up somewhat.

I'll pull the gauge and do the checks ASAP! :-)
1986 944 Turbo Garnet Rot Metallic

#5

User avatar
Tom
Site Admin
Posts: 8938
Joined: Fri Jun 25, 2021 2:04 pm
Location: Silicon Valley, CA
Has thanked: 934 times
Been thanked: 4005 times
Contact:
No movement? Hmmm... One thing you could do before pulling it out is check the resistance on the float, then compare that to what the gauge reads when the float is connected (just to see if the signal is making it to the gauge ok). You could also fill up the tank and check the float resistance to see if it's producing a full reading... As you say, could be multiple things -- float could be stuck plus the gauge could be off a bit.

#6

User avatar
Belgian951
Posts: 224
Joined: Tue Oct 03, 2023 3:59 am
Has thanked: 74 times
Been thanked: 33 times
Tom wrote: Mon Feb 10, 2025 3:53 pm No movement? Hmmm... One thing you could do before pulling it out is check the resistance on the float, then compare that to what the gauge reads when the float is connected (just to see if the signal is making it to the gauge ok). You could also fill up the tank and check the float resistance to see if it's producing a full reading... As you say, could be multiple things -- float could be stuck plus the gauge could be off a bit.
I think I'll fill up the car first and check. Thanks for the tips!
1986 944 Turbo Garnet Rot Metallic

#7

User avatar
Darwin
Posts: 317
Joined: Fri Jul 02, 2021 4:17 am
Location: Lompoc, CA
Has thanked: 116 times
Been thanked: 74 times
If you need a gauge I have one out of a spare cluster you can have. I don't need anything off of it anymore.
1984 VW Rabbit Pick-up - Not stock
1988 944 Turbo S - Really not stock (Chris White special)
2012 VW Tiguan - Kinda stock
2013 Cayenne Base - 6 Speed! Tastefully modified, mostly stock

#8

User avatar
Belgian951
Posts: 224
Joined: Tue Oct 03, 2023 3:59 am
Has thanked: 74 times
Been thanked: 33 times
Darwin wrote: Wed Feb 12, 2025 4:44 am If you need a gauge I have one out of a spare cluster you can have. I don't need anything off of it anymore.
Wow that's so generous! I'll check my gauge for issues ASAP and report back :)
1986 944 Turbo Garnet Rot Metallic

#9

User avatar
Belgian951
Posts: 224
Joined: Tue Oct 03, 2023 3:59 am
Has thanked: 74 times
Been thanked: 33 times
Update: I still haven't had the time to work on the car... It's been busy with a kid, new kid coming, a massive house renovation, new job, ... I haven't forgotten this thread though! I hope to work on the car soon.
1986 944 Turbo Garnet Rot Metallic

#10

Post Reply