chart928s4

Hi everyone, I would really appreciate some help with some electrical weirdness in my '91 S4 auto. Here is the story so far.
There were no electrical problems when I bought the car. AFAIK the pod and dash have never been removed. I installed a new stereo head unit and amplifiers a while ago. All the stereo components work well. Sometime after that - not sure if it happened immediately - I noticed the following symptoms:
With headlights on:
1. Fuse 9 is blown
2. Center console lights (HVAC, clock, etc) do not light.
3. Instrument panel lights are bright and otherwise normal
4. Instrument panel dimmer doesn't work
I removed the side console panels and examined the wiring for shorts, insulation problems, etc. and found a very small abrasion in the insulation on the black/blue wire to the ashtray light. Per @WallyP this wire is one that will cause these symptoms if shorted. I taped it up.
I then checked the resistance on fuse 9 and it was ~12 ohms, per @Alan a normal value.
Replaced the fuse, turned the key to Start, and turned on the headlights. Behavior now:
1. Dimmer apparently on the the lowest setting
2. Lights lit but extremely dim on the instrument panel. Center console lights not visible but possibly lit too low to detect.
2. Upon turning the dimmer to a higher setting, fuse 9 again blows, the panel lights go to a high brightness, and the center console lights are again out, i.e. back where i started, i.e. the BK/BL wire insulation was not the problem.

I read two long and involved threads about very similar problems but they were so convoluted I couldn't follow what the diagnostic steps. My initial instinct is that if the fuse blows when I turn the dimmer it is a bad rheostat but how to test that? If that's not the issue, is there a stepwise test process to isolate and correct the short? Thanks in advance for any and all advice.

Charlie

TheoJ

hi,
if the 58 signal is missing the cluster will not know that you have turned on the lights, so it will run full brightness until the sensor in the cluster dims it (the cluster has a little sensor to monitor ambient light).
When you turn on the headlights the 58 signal arrives on the cluster and also on the Rheostat that dims the cluster. It switches over to modulated background. The input voltage is 12v (from the lights that are switched on) and current flows through the Rheostat which is a variable resister. Due to current and the Rheostat setting the voltage drops from 12v to something like 5v. There is a series resistor in the circuit to dissipate part of the heat and lower voltage for the cluster. Check D22 on schematic. The cluster itself is part of the circuit that connects it to gnd. The voltage at the cluster pin 14 plug 4 is what manages the modulation of the backlight bulbs in the cluster. Your problem may be somewhere inside the cluster or just before.
One way of telling is to disconnect the #4 plug, make sure the fuse is fine, and turn on lights. Turn the Rheo up and down. If the fuse blows you have to check wiring, Rheo, and resistor. If it doesn't, connect a 5w bulb to the #14 pin of the connector and gnd. See if this still keeps things working. If all is ok, the cluster comes into play I'm afraid.
regards, Theo
928gts.jenniskens.eu

chart928s4

Thanks Theo. I will check those items and hopefully declare victory or at least report back. From your suggestion it does not appear that something easy like "replace the rheostat" is going to work. Sigh.

BTW - thanks also for all the work on your website with its tons of great info. I haven't seen you posting here lately but I'm glad you picked this post to jump in.

chart928s4

Reporting back...

In an effort to avoid removing the pod I decided to gather some data (daylight test so sensor light control is probably negligible):

As a baseline check, I blew a few fuses to see what if anything caused the fuse to blow. At one point, the fuse was ok until I turned the rheostat, then it blew. So I decided that would be the first thing to check.

1. I removed the "potentiometer"/rheostat.
- The rheostat bench test at the lowest resistance was .5 ohm; at the highest 5.5 ohms.
- With the removed rheostat, key to ON, and headlights turned off then on, the fuse does not blow
- Using a Fuse Buddy with a 10 amp fuse in Fuse 9, current was negligible
- The cigar lighter light remains on even with the key off so I removed the bulb


Fuse Buddy fuse amperage meter

Cigar lighter light

Melted!

2. With the rheostat replaced, the key turned to ON, and the headlights cycled on
- The Fuse Buddy registered 5.5 amps
- Within seconds the dangling rheostat started smoking (quickly turned the key to OFF and yanked the Fuse Buddy)

3. Removed the rheostat again and installed a new 5 amp fuse. With the key turned to ON and the headlights cycled on, the fuse doesn't blow. Instrument panel backlights are visible.

At first glance it seems like the rheostat was bad (well, it's definitely bad now). However, it's a pretty simple device and there was no obvious problem when I took it apart. I would prefer not to fry a new one because of an upstream problem. The cigar lighter light is mysterious. It is fed directly by Fuse 9 which is fed by the headlight switch if I read the diagrams correctly. Could it mean a problem with that switch? Finally, where is the "resistor instrument panel" located?

chart928s4

Update: In the dark, with the rheostat uninstalled, the instrument backlights are on. With the headlights cycled on, the backlights go out. Seems like expected behaviior? If the headlights are off, the power comes direct to the panel? Headlights on, governed by the rheostat?