Date: Thu, 11 Jan 2001 21:22:23 -0500
Reply-To: David Beierl <dbeierl@ATTGLOBAL.NET>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: David Beierl <dbeierl@ATTGLOBAL.NET>
Subject: Re: Help! Appeal for data
In-Reply-To: <3A5E21E3.7715998B@earthlink.net>
Content-Type: text/plain; charset="us-ascii"; format=flowed
At 04:13 PM 1/11/2001, Mark Drillock wrote:
>David, this variation has bothered me for years and I hope you can get
>to the bottom of this. I have seen both behaviors in several Vanagons
>but it never made sense as the wiring diagrams each look pretty much the
>same in this area.
I've looked at the diagrams for all years and they're electrically the
same. They move the wires around a bit but the bottom line is that (on the
diagrams) the gauge/blinker has only three terminals. One goes to the 10v
regulator, one goes to ground, and one goes to the sender which is wired in
parallel with the output of the low-coolant warning circuit. The two
gauges I have here (from '84 plus one from a panel that had a green LED for
high-beam) appear identical, with three terminals. An electrolytic cap is
visible inside the case; one has an aluminum and the other a
tantalum. With ten volts input to the gauge, the light starts flashing
when the sender terminal is drawn down to 4.8 volts, which puts the needle
just touching the bar at the top of the scale. There may be some
interaction (in the millivolt range) between the needle position and the
light flashing, but it's very definitely a second- or third-order effect.
>It seems like the possible explanations are a wiring error, a defect in
>the warning control unit, or a design difference in some units. Have you
>tried removing the level warning control unit to see if that disables
>the level warning? What is the part number on your warning unit? What is
>inside the thing, do you know?
The controller I'm looking at (which comes from the '85 GL by the way -- it
donated a panel with tach to the '84, but not the controller) is a
251-919-376, number 43 stamped in white on the top, comes from the hidden
relay panel above and forward of the regular fuse/relay panel -- very easy
to get to with the dash removed! (when they changed to the new fuse panel
they moved this onto the much-expanded relay section of that panel). It
has terminals 15 (+), 31 (Gnd), S (Sender), G (Gauge) -- at some point they
shifted to numbered terminals instead of G and S. It contains a 4001, a
transistor (case like a 2n3906, I forget what it's called) which feeds the
output -- open collector I think; couple of electrolytics, two big and two
signal diodes, three ceramic wafers, and eleven resistors. When hooked to
power and gauge, if the S terminal is connected to either B+ or ground
nothing happens; but if the S terminal is left open, after about five
seconds it drags the G terminal down to about 3.5v (1.9v open circuit)
which immediately starts the light flashing and the gauge
climbing. Several seconds after the S terminal is tied again, the gauge
starts dropping and a few seconds later the light stops flashing.
All this appears to be doing exactly what it was meant to do...
>My 83.5 would blink the led with a low level in the
>tank but the gauge would stay put.
It seems to me that that this *could not happen* with the wiring as
described. Can you think of a way to influence the light and not the
needle, given only one input terminal to the gauge? All I can think of is
superimposing some sort of AC signal on the DC...seems far-fetched somehow.
Thoughts? ISTM we need to find one that behaves the other way and trace
the circuit etc and find out what it's really doing...
david
David Beierl - Providence, RI
http://pws.prserv.net/synergy/Vanagon/
'84 Westy "Dutiful Passage"
'85 GL "Poor Relation"
| 
