A. D. Smith (a.d.smith@umds.ac.uk) searched for insight as follows:
>The one that I think is most likely (or at least the one that I want to be
>true) is that the problem lies with the "Bimetalic voltage stabilizer" in
>series with the gauge / sender. The funny thing is that I replaced this
>about a year ago !
This may be a clue. Was it really dead? Do you still have the
"broken" one? What is the resistance across the "new" one versus
the old "broken" one?
>Does anyone know exactly what's in there, so that I can bring it
>to work and get the electronics guys to test it ?
The "bi-metalic" in the name indicates that two different metals
(for example, copper and tin) are used to create a variable
resistance. The current running through the device creates
heat, which causes the two different metals to expand at
different rates, and bend a metal tab. The bending tab "shorts
out" a resistor at different points, thus making a variable
resistance. If you want a "bench test", measure the device's
resistance with the ohms setting on a volt meter. It should
conduct electricity (and display a low resistance reading).
Testing the response of the stabilizer under different "load"
conditions would require all sorts of math and a few rigged-up
test circuits. Forget about getting that detailed.
The operative question is "Do you have 12 Volts DC (or something
more than 0 Volts DC)? You would want to put a decent volt meter
on various points WHEN EVERYTHING IS IN CIRCUIT IN THE CAR.
Better that you borrow a volt meter from your pals than bring the
voltage stabilizer into the lab for them to "test".
In short, the stablizer is a variable resistor, and so is
the fuel-tank-sensor. The fuel gauge reading varies with
the changing resistance of the fuel-tank-sensor, so...
A) If the gauge reads "Empty" when you have gas in the tank:
- The tank-sensor wire could be unplugged or broken
and somehow grounded, most likely because it is
touching metal somewhere.
- The fuel gauge could not be grounded well (it should
have its own little black wire attached under
one of the knurled nuts that hold the gauge in
place, and the connection should be tight and clean.)
- The gauge itself could be bad.
- There could no no voltage fed to the circuit, for
whatever reason.
B) If the gauge reads "Full" all the time:
- The wire from the tank-sensor to the gauge
is broken or unplugged, but the good news
is that you DO have voltage.
Not knowing the specific symptoms of your problem, I would suggest
the following test process if the hints above do not help:
1) If you can borrow a volt meter from your lab, start at the fuse
box. Look for the fuse with the multiple green wires connected.
Measure your voltage there. It should be roughly 12 VDC with
the key turned "on". If it is less than 12 VDC, compare the
reading with the reading across the battery terminals. The
two should match.
2) Pull the cable connector with the multiple green wires off its
terminal. This should disconnect all of the following:
Emergency Flashers
Reverse Lamp
Stopping Lamps
Fuel Gauge
Tachometer
Whatever else you have kludged onto those circuits.
3) Make up a cable (get some push-on terminals at your auto
supply store, and borrow a crimper, or just use a pair of
pliers) that will run from the "multiple-green" fuse terminal
to the "B" connector on the voltage stabilizer. (Unplug the
green wire, as it is no longer connected to the fuse.) This
"test" wire will feed power to the fuel gauge circuit, leaving
the other items out of the circuit, and hence, making the
debugging process less complex.
4) If the problem just went away, this means that the source
of your problem is perhaps the wire itself, but more likely
the problem is caused by one of the other circuits mentioned
above. (Reconnect everything and see what works and what does
not work, or trace each circuit in turn using methods similar
to those outlined below.)
5) If the problem persists, unplug the "test" wire from the "B"
terminal of the voltage stabilizer, change the connector on
the end of the "test" wire from female to male, and plug it
into the Light Green/Green striped wire that connects to the
"I" terminal of the voltage stabilizer. You have now removed
the voltage stabilizer from the circuit, and are feeding the
tank sender and gauge with a direct (raw) 12 VDC feed from the
fuse. If this clears up the problem, perhaps you DO have a
second failure of the stabilizer, but I have never seen a
dead one in 20 years of messing about with MGs.
6) There are a couple of bullet connectors in the circuit that
are not shown in the wiring diagrams. They are located
at the front edge of the fuel tank, and they are a failure-
prone aspect of MG Midgets. One wire (the Green/Blue) is
the connector to the fuel tank sender unit, and the others
are the power feed for the fuel pump. All can be looked
at if you put the car up on jackstands, and all should
be inspected. (I have soldered mine, since I hate the
idea of a bullet connector that is exposed to random
junk being tossed up from the wheels and road.)
7) A bad ground at the fuel tank can be a real pain too.
The fuel-tank-sensor has ONE wire connected to it. This
is the "hot" wire to the fuel gauge. The tank-sensor
gets its ground connection from the metal-to-metal contact
between the sender and the fuel tank (that funny-looking
round twist-lock mounting system). Some emery paper
and elbow grease applied to both surfaces might help.
The tank gets its ground from the studs to which one
attaches the nuts that hold the fuel tank. Select one
of these nuts, remove it, clean the stud, replace the
nut and washer with new ones, and emery-paper the heck
out the the tank flange in the area under and around the
nut and washer. (I have a specific braided copper ground
strap on my fuel tank, and therefore have eliminated
this "problem-just-waiting-to-happen" on my MG).
8) The meters themselves are not very reliable, nor accurate.
Current flow through a hand-wrapped coil of wire causes
the motion of the meter, and they are subject to mechanical
vibration, even under the best of conditions.
9) If you have good connections, and still have no luck,
measure the voltage at each possible connection and report
back. I can measure the voltages on my (working) Midget
to compare. If you have the MG shop manual or the Haynes
manual, this would help, since they number the components
in their electrical schematics.
10) I have heard of (but never seen) strange things happening
with Midgets that have a bad ground connection at the
ground strap near the clutch slave cylinder. This is the
"main" ground for the entire car, and it can become coated
with oil, clutch fluid, and grime to the point that there
is no longer a ground. Some cleaning here might be in
order if all else fails.
>While I'm here, does anyone have any suggestions about clocks...
Funny you should mention that. I made a "rally" instrument panel
for my Midget that replaces the metal-mesh panel in the center
console of a Midget. I took a collection of VDO gauges, and did
some painstaking work with the meter faces to make them look like
Smiths gauges. I then had the face rings chromed to match the MG
"style". I have:
Volts (Same Size as Fuel Gauge)
Amps (1.5")
Block Surface Temperature (thermocouple-driven) (1.5")
Vacumn (Same Size as Fuel Gauge)
If I wanted, I could pop out the "rally" panel, put the chrome
mesh back in and put the stupid little light back on, but I like
the instruments better. This MG is not for show, it is for
driving, so I have no qualms about enhancing the car far beyond
what British Leyland had in mind.
But a CLOCK? Whatever for? Time is relative, and if I am in
my MG, I will be either very very early (driving too fast) or
very very late (taking the scenic route). If I had a clock, I
would look at it, and think about what I was supposed to be
doing, thus making the MG just about as much fun as my Volvo!
The 2nd Law of Thermodynamics says "Natural processes increase entropy".
Translation - "You think you're in trouble now? Just you wait..."
james fischer jfischer@supercollider.com
|