Rheostat input -> Voltage output

Assuming you can't rewire the rheostat to "hang" from +12V, you have an upside-down problem. If you want to toss in a cheap op-amp (LF411?), that'll do it with just a few extra resistors.

Tim

-- Deep Fryer: a very philosophical monk. Website:

I'll post a schematic later today -- got a "honey-do" to do right now. A dual LM358 opamp (8-pin DIP, RadioShack) and a few resistors will do the job.

OK - pretty sure that the old gauge would have been doing the damping and so the adaptor gadget would need to damp the output from the rheostat.

I guess this could be done by increasing the interval between samples from the rheostat??

Thanks Don, that looks great!

Is That Amp is the LM358 you mentioned above?

Also, on the LM317, which is the "output voltage" and which is "adjust"? (input voltage is pretty obvious :-)

Schematic at

If you need damping, add an RC section with a voltage follower (made with unused amp) between sender and the 30K resistor. 100K series and 10uF shunt to ground would give a 1-sec time constant, etc.

A voltage follower is just an amp with the input going to the + input and the - input connected to the output.

I have a sender here out of the tank and looking at the rheostat, it is very non-linear, but I think that is to allow for the non-linear nature of the tank since the spare tyre is in the top 2/3 or 3/4 of the tank

Since the circuit is so simple and parts should be very cheap, I think I'll knock one up and test the whole setup on the bench with the spare sender and a spare dash module

I'd say that you would be right to say that the new sender is a linear device since the spare in those cars is beside the tank

It depends on whether the sender or gauge does the damping. If the sender is driving a heat actuated gauge then the gauge damps. But it could be that the sender is driving something faster and then the sender must be damped. See

as an example. You will be wasting your time building anything until you find out what is what with this sender.

Test

A word of caution before you start building the hardware. The fuel gauge senders that I remember from cars in that era were very non-linear, but that was matched by the nonlinearity of the dash meter. If your 1986 dash meter is a linear meter, then you obviously have problems. Your tank may only be

10% full, but the dash gauge tells you that you have half a tank. Not really good on a long stretch of road with no gas stations. You should probably draw a graph of the sender's response before you start. Start with an empty tank (drain it with a siphon), then add a gallon of gas, measure the resistance. Add another gallon, measure the resistance. Etc. Etc. Also, measure the response of the dash gauge. Get a power supply,a variable resistor and a multimeter. You'll have to experiment with the value of the resistor, since you probably don't know the sensitivity of the gauge. Make a graph of the current vs. readings of the gauge. If the gauge is linear and the sender is nonlinear, then you have a problem with the circuit that was recommended, in that it will be grossly inaccurate with the original sender. Cheers!!!

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in 
the address)

Never take a laxative and a sleeping pill at the same time!!

Do ya' think? It depends on the tank. If it is stepped and contoured, and the gauge is just a linear height reading in a well, then the external circuit has to fit the tank volume to the well height resistance function. This can be done with a bimetal, external resistor network, and/or the method of float. It is very unlikely the straight resistance to voltage conversion will work with a linear gauge. This is sounding more like a job for a PIC.

Yes, the amp is LM358 (dual) or LM324 (quad).

Forgot to label the 317, sorry 'bout that. BATTERY goes to "in", OUT goes to the top of the 24 ohm resistor, ADJ to the bottom of the resistor. The regulator maintains a constant 1.2 volts across the resistor, resulting in a constant current of 25 mA thru the resistor and thence to the fuelgage variable resistance. This results in a voltage that is directly proportional to fuelgage resistance. Said voltage is then inverted, translated (because resistance doesn't go to

0) and scaled by the amp to provide the desired output.

Damping can be provided by an intermediate or following stage with DC gain of 1 and a single RC timeconstant as desired. _________

----R---------| + | | | | C | amp |________________ | __| - | | | | | | | | | |_________| | | |_______________ | | gnd

This cap may be electrolytic, + up. Timeconstant will be R*C. I'd keep R below 1meg with electrolytic caps. A tantalum cap might be best here, for low leakage across temperature.

Right you are, thanks. Should be 49.6 ohms. Use 51 ohms as nearest

5% standard value or 49.9 ohms as nearest 1% value, or 68 ohms paralleled with 180 ohms to nominally make 49.35 ohms.

That circuit is right off the specsheet in the applications section, is commonly used as a current source.

As you noted and I responded, the resistor value should be changed.

How it works: the LM317L regulates a series pass transistor from INPUT to OUTPUT terminals, passing enough current to maintain 1.23 volts from OUT to ADJ in presence of whatever external circuitry is there. In this case, it maintains 1.23 volts across the previously-24ohm, now-49 ohm resistor. The voltage on the OUT terminal will be whatever it must be (within operating range) to maintain this condition. Constant drop across a series resistance

---> constant current thru resistance and (varying) load to gnd. This will certainly be true for loads of 13 ohms (0.325 volts from ADJ to ground or 1.55 volts from OUT to gnd) ) to 100 ohms which will be

2.5 volts from ADJ to ground or 3.373 volts from OUT to gnd -- well within the operating range of the 317 with 14.4 volts from INPUT to gnd.

That circuit looks dangerous as drawn- by the usual interpretation he has that 24 ohm in series with ADJ and OUT directly to the load. The LM317 will just saturate out, and on the full tank setting this would be somewhere around 0.8A. You would want to do something like this: View in a fixed-width font such as Courier.

. . . LM317 . +---------+ . | | 9.8V . BATT--+--|IN OUT|----+ . | | ADJ | | . | +---------+ | . | | [120] . === | | . |0.1 | | . | +---------+ . --- | . /// | . [820] 1/2W . | . +--------->Vout . | . | . | 0.13v->1.0v . | . /_ F E . \\/| . / . /\\ . | . | . --- . /// . . . . 1.0. \\ . . \\ . . \\ . . \\ . . \\ . . ---\\ . . \\ . . -- \\ . .13. \\___ . .................. . . E | | | F . 1/4 1/2 3/4

Revise that: View in a fixed-width font such as Courier.

. . LM317 . +---------+ F E . | | 8.5V -> 9.4V . BATT--+--|IN OUT|----+ . | | ADJ | | . | +---------+ | . | | [120] . === | | . |0.1 | | . | +---------+ . --- | . /// | . [680] 1/2W . | . +--------->Vout . | . | . | 0.13v->1.0v . | . /_ F E . \\/| . / . /\\ . | . | . --- . /// . . . . . 9.4 1.0. \\ . . \\ . . \\ . . \\ . . \\ . . ---\\ . . \\ . . -- \\ . 8.5 .13. \\___ . ....................... . . E | | | F . 1/4 1/2 3/4

Your having a problem with Ohm's Law today because 1.24V/24Ohms=52mA.

So add 240 ohms from ADJ to the sense resistor/amp-input node. That would limit current to 56.8 mA worst case if the 317 shorts, would would have no effect in a properly operating circuit. There's plenty of voltage headroom for the 317 to still work properly with ADJ at 8.5 volts (when Rgage = 100 ohms, vgage = 2.5 volts) during normal operation. This has the additional benefit of reducing the dissipation in the 317L to about 170 mW max.