Today's progress.  Speaking of glow plugs, it is possible to work
on and modify the old series-plug relay.

Replaced the horns.  For whatever reason, both old ones were dead.
The vacuum tank had leaked down by morning, I don't suppose that is a

I next tackled the glow plug relay, which has been having some
intermittent faults.  (No light.)  I opened the harness connector and
squeezed the sockets together a bit (to tighten them up) and hit them
with contact cleaner.  I then burnished the pins on the relay and
opened it up to check the capacitor values.  (It's interesting that
the current to the plugs is sensed via a 1-turn transformer around a
reed switch.)  Two of the larger caps were bad, so I replaced them.
The big one (2200uF) that runs the light was OK, oddly enough.
Unfortunately while messing around on the bench I managed to slip with
a probe and blew up two diodes, one a 6.8V Zener.  This took some time
to find!  I had a junkbox 6.2V Zener, I call that close enough.
Anyway, after several hours of circuit tracing and head scratching I
got it back to where it was before I killed it.  (We won't be
'billing' for that time!)  The beauty of electronics from this age is
that there are rarely special parts inside, component-level repair is
entirely possible.

Next, check the timing.  I want to try to extend it to the long side
to make up for the engine's age.  It obviously needs a lot of preglow!
I dug some numbers out of the MB manual:

        deg C   deg F   Ohms    Spec.   Measrd. Modified
        -----   -----   -----   ------- ------- --------
        -30     -22             45-65
        -20     -4              35-54
        -10     14              26-46
        -5      23              30 nom.
        0       32      8500    20-35   27      40
        20      68              12 nom.
        25      77      2500    8-16    9       15
        80      176     300     0-4     0.5     1

On the bench with the resistance at 8500 Ohms the time was 27
seconds.  The appropriate trim pot inside could make the light's time
shorter, but not any longer, so I paralleled an additional 1000uF
capacitor with the 2200uF one, which extended the light time to 40
seconds.  (All that circuit tracing helped decide how best to do
this.)  The glow timeout is about 150 seconds.

Anyway, after all this the time was better but the light was still
flakey on a test drive.  (I was picking up chainsaw chains, and the
central locking worked fine when I tried it.)  Back on the bench it
was sensitive to twisting and bending.  So though I had resoldered
almost everything already I had not done the heavy relay and connector
pins, which are the most likely candidates as they are so big (cold).
I did those and then it was no longer sensitive to flexing.  Back in
the car the light behaved as I'd expect.  Whew!

-- Jim

Reply via email to