Fuel gauge question - electronic guru needed

[ischmitz]

I have swapped all five gauges in the 74 to 993 gauges. I gave me correct readings of the temperature with the 3.6, a working clock, no more cracking noises of the plexi glass, and much nicer looks overall IMHO.

However, there is one question for the electronic gurus amongst you guys:

The 74 stock fuel level sender is a simple passive variable resistor that varies between 4 Ohm (full) and 68 Ohm (empty). It is hooked up to GND and one end of a simple Ampere meter (the stock gauge). Current flow like this:

GND - var. resistor - gauge - +12V.

The 993 sending unit (which won't fit into my gas tank) for the newer 993 gauge is a semiconductor device. After having done some tests with the new gauge it seems to read between 0mA (empty) and 10mA (full). Does anyone happen to have confirmed specs on this?

Just hooking up the old sending unit to the new gauge leaves the gauge pegged at full all the time. Current is about 50mA. Too bad.

I guess I need something like an OpAmp based voltage to current converter and make the input a voltage divider with the old sending unit being the variable resistor. Has anyone ever done something like that?

Ingo

[HawgRyder]

Or...you could just put a resistor in series with the power.
The idea being to reduce the current swing down to the 993 level.
If 10ma is the max for the guage....then select a resistor that will add to the resistance of the sender to equal the resistance of the 993 sender.
If it draws 10ma with 12V....then ohm's law says 12V / 1200 ohm = 10ma.
Measure the resistance of the original sender...and add the correct amount to the circuit.
The correct order should be:
Gnd - Var Res (sender) - Guage - added resistor - 12V.
You want the sender to be able to ground the circuit.
Good luck
Bob

[Joeaksa]

Ingo,

How about putting a variable resistor (a pot) inline and vary the setting until you get a valid empty reading, then move the float up and confirm that it shows full. Measure the resistance through the pot and get a fixed resistor of that value. I would make it accurate at the lower level over upper if both did not work out at the same time.

Joe

[ischmitz]

Bob,

that's what I thought about initially, too. But in this case the gauge would see something like 1268 Ohm (full) and 1204 Ohm (empty).

In other words, I'm afraid it's not that simple since the dynamic range with this solutions is next to nothing.

Ingo

[thabaer]

You'll probably end up with linearity problems, but maybe a divider AND a current limiter ?

A handful of pots and you might be able to just trial and error up the right mix to get a reasonable empty/half/full scale. Maybe dig up an op-amp cookbook if you can't get a win from the resistor network.

Can you get the sender out to do the fiddling on a bench w/ a 13V supply ?

I just realized that the 993 probably had a VR feeding the original cluster (and that you already mentioned opamps )