[T3] Fuel gauge issues (jumpy needle)

[Jim Adney]

On 31 May 2015 at 8:01, Jacob Adam Schroeder wrote:

> That black wire, for me, is soldered to the black plastic pod.  Is that
> really how it is supposed to be connected or is it supposed to be a push-on
> style connector?  Either way, that is more a curiosity than related to my
> present issue.

Yes, that black wire goes inside the black lamp base and solders to 
one of the circuits there. This is rather clearly shown in the wiring 
diagrams in the Bentley. Not so clear on those wiring diagrams is the 
brown ground wire: That's shown as if it's connected to the metal 
gauge body, but no such connection exists. It actually connects to 
tab #31, as you describe below.

> > There are 2 grounds in the circuit:
> >
> > 1 is at the  gas tank, where the sender is grounded via its mounting
> > bolts. The sender does not need an unpainted place there, as the
> > bolts should make contact in the threads in the tank. You could turn
> > those bolts back and forth a bit to make sure that they have cut thru
> > the paint. The tank also needs a good ground to the body, so make
> > sure the 4 clamps and bolts around the edge of the tank are tight.
> > You could loosen and retighten each of those to make sure they are
> > making a good connection.
>
> Because I recently repainted the exterior of the tank, I will give this a
> shot.  I can see how the threads in the sending unit bolts could make a
> good ground connection to the tank.  I may turn those back and forth a bit
> to make sure any paint is cleared out.  But what about the paint along the
> edge of the gas tank where the 4 clamps are?  Those clamps (IIRC) thread
> into the body and only press against the exterior of the tank.

If you had done something to protect the paint from those clamps, 
like put a  plastic sheet in there to prevent paint damage, then you 
would have a  problem. Otherwise, it seems unlikely that at least one 
of the clamps would have cut thru the paint.  

Either way, it's easy to test this with an Ohmmeter. Just check for 
continuity between the sender base and the body. You should get 
something less than an Ohm. For the least confusing results, 
disconnect the wire from the sender while you check this ground.

> > The second ground is via the spring clips that hold the gauge in the
> > dash. Sometimes these get spread too far out and they loose tension.
> > There's a hump in the dash that those spring clips should NOT go
> > outside of. Or sometimes it's just that the plastic dash has shrunk
> > slightly and no longer hold the spring clips tight against the
> > backside of the dash.
> >
> > If the gauge head ground is bad, the gauge needle will jump around if
> > you turn on the 4-way flashers. I'm not sure if this happens with the
> > key OFF or ON, or both. With a good gauge head ground I think it will
> > not happen at all.
> >
> I do get some (very minor) jumping in the fuel gauge needle when using the
> 4-way flashers.

I THINK this indicates that there's a poor ground somewhere. 

As a test, I just went outside and checked 2 cars in my driveway. The 
'69 gauge needle did not flicker at all with the 4-way flashers, 
either with or without the key ON. It's tank was nearly empty, 
meaning <10 Ohms in the sender.

BUT, on my '73, the gauge flickered wildly with the key OFF, not at 
all with the key ON. However, this may have been due to the tank 
being full, meaning ~100 Ohms in the sender. I haven't noticed 
flickering gauge readings on that car, but it looks like I need to 
look into the grounding.  

> > It should be possible to add a 3rd ground in the form of a brown
> > (ground) wire from an appropriate place in that plastic piece out to
> > a ground terminal under the dash. This will fix a poor ground at the
> > dash.

> I think I may have this ground already.  My black plastic piece has a
> terminal (labeled 31) with a brown wire connected to it that appears to be
> ground.  Once it connects to the black plastic piece, it is connected to
> the gauge pod through one of those copper nipples on the back of the gauge
> pod.  Assuming this provides ground to the body of the gauge pod, then why
> would the ground through the spring clips be necessary as opposed to
> redundant?  Are the grounds for the lights separate than ground for the
> fuel gauge? 

Okay, you've clearly looked into this carefully, so I went down and  
looked at several of those plastic bulb bases. You're right, they 
have a term 31 on them. Yes, that's ground. I had forgotten that 
terminal was there. Yes, from '69-on there should be a brown wire 
that daisy-chains all the dash instruments together, including the 
gas gauge, and then connects to a tab on one of the braces under the 
dash. Yes, that ground makes the spring clip ground redundant, so I 
think you should follow that brown wire and make sure that it's 
getting to chassis ground properly.  

On cars earlier than '69 that brown wire would not exist, so either 
that #31 tab was left bare, or the early black lamp bases had only a 
rivet there and no tab. Or maybe someone with a '68 or earlier can 
fill in the details.

I think you're on the right track, but I don't know anything special 
to suggest. One thing to keep in mind is that it's possible for there 
to be a poor connection between that #31 tab and its rivet. One more 
data point that I remember from long ago was that my very late '71 
had a bad problem with that ground, and it was fixed by getting a 
better ground at the spring tabs. Maybe this points to a systematic 
problem with that #31 tab and its rivet....

Another test one could do, would be to do the 4-way flasher test at 
night with the instrument lights on, and see if the instrument lamps 
also flickered with the flashers. That would tell you which side of 
the ground wire the bad connection was on. I'll try to remember to do 
that test on my '73 tonight.

The actual gas gauge circuit is a bit of a mystery to me. If you take 
it apart, 
you'll find 2 different resistors wired in with the meter movement in 
a way that doesn't lead to any obvious conclusion. Since it has 
always worked for me, I've been willing to give VDO credit for a good 
design and move on to more  pressing issues. If you manage to figure 
it all out, please let us know how it actually works.

-- 
*******************************
Jim Adney, jadney at vwtype3.org
Madison, Wisconsin, USA
*******************************