Ok, having slept on it I have some new ideas. If I can repair the
cluster rather than replacing it, I'm willing to do more diagnosis  --
seems like replacing a cluster and ending up with a new ODO, temp and
fuel gauges with possibly different calibrations is not a trivial
matter.   Have a soldering gun and am prepared to use it!

* Clue 1: Shorting across 30/58b, instead of lighting the instrument
lights, causes the LEDs to glow.  With engine running, it also causes
tach, fuel & temp gauges and clock to go off.

* Clue 2: Instead of shorting across 30/58b, if I put the switch back
in and use it, I find that with headlights on (which gives panel light
power) the panel dimmer rheostat can vary this behavior :  if it's at
the dimmest setting, the gauges work, but as I brighten it up, they
all start dropping and finally go off (as in clue 1).

* Clue 3: My ohm-meter, touching between the car metal and T14/3
(ground connector) reads 0 (well, really 0.2 but that's what my ohm
meter reads at zero).  I also wiggled & unplugged & replugged every
connector on the two "star" connectors.    My conclusion is that it's
not a ground problem for that side of the circuit.

Trying to reason this out using the old V=IR...

Theory 1: could an blown instrument bulb cause this?  I don't think
so, because the 4 bulbs are in parallel, so one dead one should not
have an effect.

Theory 2: could a shorted instrument bulb cause this?  These are tiny
bulbs so they should have very high resistance, so normally the ground-
side of the instrument panel will be at zero volts or close to it.
However, if a bulb shorted, then the ground side of the circuit could
float high, but only if that ground circuit is dirty.  Since T14/3
seems well grounded to me, this suggests a short in the ground circuit
in the cluster itself?

Theory 3: one thing that doesn't fit -- the failure mode involves the
high beam LED lighting up, which is illogical: the "high floating
ground" theory would REDUCE, rather than increase, the potential
across that LED.     I notice the Voltage Stabilizer on the diagram
right nearby -- would a failing voltage stabilizer cause this?

Theory 4: faulty voltage regulator?  that looks easy to test, will go
test it.

On Dec 22, 2008, at 4:57 PM, mdrillock wrote:

> That ground 16 is explained on 97.66. Look at the bottom of the page
> and
> current track 104.
>
> I have written about the dash ground harness weakness various times
> over
> the years. VW connected too many things to a single ground point and
> if
> that single point gets bad all kinds of screwy things happen. Best to
> isolate the cluster ground with it's own wire tied to a screw. Then
> the
> other dash stuff will not feed back into the instrument cluster.
>
> You do have a 1985 with a 1985 cluster I hope?
>
> Mark
>
> Michael Diehr wrote:
>> Ok, more info:
>>
>> I'm 99% sure it's not the headlight switch, as I've isolated the
>> problem : when I short across connector 30 to 58b (instrument panel
>> lite dim), instead of getting instrument panel lights, some of the
>> LEDs light up.  If I do this while the engine is running, then the
>> tach and clock go dead.
>>
>> According to Bentley page 97.63, this current track goes as follows:
>>
>> terminal 30 on headlight switch to terminal 58b on headlight switch
>> via the dimmer
>> Fuse S13
>> Track 68 (cigarette lighter light)
>> Track 67 (blower fan control lever light)
>> (to a mysterious circle labeled 22)
>> to the instrument panel lights via T14 connector 1
>> out to instrument panel ground via T14 connector 3
>> To ground, labeled 16
>>
>> If I had to guess, I'd guess that Ground #16 is bad.  However, there
>> is no ground #16 listed in benteley on page 97.60.
>>
>> I checked and cleaned the other grounds (including ground #10 the
>> star
>> connector above the fuse box -- however my van only has one star
>> connector there).
>>
>> Any idea where ground "16" lives?
>>
>>
>>
>> On Dec 22, 2008, at 11:14 AM, Ken Wilford wrote:
>>
>>> It does sound like a ground issue to me as well.  Check the grounds
>>> that
>>> are located behind the fuse panel.  You should be able to remove a
>>> couple of screws and drop the panel down and you will see the ground
>>> connections to the left.  There are a bunch and they are clustered
>>> together on a grounding "star" connector.  Check each ground on
>>> these
>>> two connectors and clean them as you go.  You will probably find
>>> that
>>> one of the ends of the ground wires has broken and you will need to
>>> crimp a new one on.  That is what I believe is going on.  The ground
>>> has
>>> gone bad and the headlights are seeking a ground through the
>>> instrument
>>> cluster.  You shouldn't need a new headlight switch but if you find
>>> you
>>> need something please let me know.
>>>
>>> Ken Wilford
>>> John 3:16
>>> www.vanagain.com
>>>
>>>
>>> Michael Diehr wrote:
>>>> 1985 Westy -- just last night started misbehaving.   Symptoms:
>>>> with
>>>> ignition off + headlight switch on, three LEDs in the dashboard
>>>> glow.
>>>> With engine running + headlight switch off - normal.   Engine
>>>> running
>>>> + headlight switch on, all dash instruments (including lighting) go
>>>> dead.   The load reduction relay seems to work fine (e.g.
>>>> headlights
>>>> on + ignition on = headlights work, and when you crank the starter
>>>> the
>>>> headlights go out).  Toggling high beams doesn't do anything weird.
>>>> This sounds like a ground issue from a failing headlight switch?
>>>> Wiggling the ignition key doesn't seem to change anything in any
>>>> state.
>>>>
>>>> Unfortunately the van is about to go on a long trip, so need the
>>>> part
>>>> asap etc.  Are there any surefire diagnostics I can do?  I have
>>>> voltmeter & bentley.  Thanks!
>>>>
>>
>>