How to reverse resistance?

I have a (fuel level) sender that changes resistance from 0 Ohm (empty) to

90 Ohm (full). Fuel gauge, however, expects 0 Ohms on Full and about 250 Ohms when empty. Gauge is controlled electronically (via onboard computer), so I can't rewire the gauge itself, it has to be the sender. Any ideas on how to wire the sender?

TIA, Peter

Reply to
Peter
Loading thread data ...

Do you have the sender on the bench? Can you describe the sender?

Reply to
Tom Biasi

Sender is integrated into the multivalve design (it's liquified propane gas tank), and essentially cannot be modified...

Peter

Reply to
Peter

Well you can't modify the sender and you can't modify the receiver, the only thing left is a translator between the two. Just off the top of my head maybe converting the resistance to voltage and then the voltage to your desired resistance. Tom

Reply to
Tom Biasi

In response to what Tom Biasi posted in news:WGcXe.4126$ snipped-for-privacy@fe11.lga:

Controlled resistances are possible, but far from easy. Controlled current sinks, on the other hand, are not difficult - particularly if they can be ground-referenced.

--
Joe Soap.
JUNK is stuff that you keep for 20 years,
 Click to see the full signature
Reply to
Joe Soap

Hi, Peter -

I had a similar situation applying an automotive fuel gauge meter and sender to a boat many years ago. I eventually wound up with the following combination (view in fixed-width font):

| .-. | |R1 | | '-' R2 | ___ + +----+|___|-+---| | + +----+ | | | | .-. | ' sender | |

Reply to
John - kd5yi

Peter, you can use a simple FET based inverter, but getting it to track linearly over a wide range may be an issue.

Harry C.

Reply to
hhc314

Can you measure the current through a 250 ohm pot installed in place of the sensor? My thought is to control the current into a current mirror with an inverter in the collector leg and the 90 ohm sensor controlling it. At this point, my thought is just a wild-assed guess.

Ed

Reply to
ehsjr

It may be easier than you think. Put a 100 ohm rheostat in place of the fuel sender and vary it from full to empty and record the voltage change going into the "computer."

Inside the computer there's probably a regulated supply with a series resistor to the fuel sender. The change in voltage at the juncture is probably what matters to the computer.

You just need to drive your new sender from a regulated supply with its own series resistance and get an output voltage swing. From there it is op amps to massage the signal to change the direction of swing and get it into the range the computer needs to see.

You may need an op amp capable of driving some current on the output if they use a low value series resistance on the sender (from the computer) You can get an idea there from the current through the rheostat.

The high current op amp could probably be implemented with a Mosfet driven with a low current op amp if the output voltage on the MOSFET is part of the overall feedback loop in your op amp circuit.

It sounds easy. It would take some measurements for voltage and current and some think time, and a bread boarded circuit to try.

I don't expect linearity would be a problem if you treat it as a voltage you have to match.

(or a current) - this sort of thing is done all the time in industrial process control. The sender is a V to I sender. Takes a voltage like

1 to 5 volts DC and outputs a variable resistance to send 4 to 20 milliamps to the control circuits (that's a standard for process control - but 4-20 may not be a standard for automotive use)

Fuel senders aren't always linear. Most of the gasoline senders use a float on the end of an arm that turns a rheostat inside the gas tank. The output of the sender is a sine function (moves faster at the midpoint than at the extremes)

Or you can forget about the built in fuel gauge and design your own around the sender you have. Run a 10 or 20 segment bargraph LED display for fuel level . . . Easier and less hassle.

Car fuel displays are very slow to respond (heavily damped) to account for uneven terrain and fuel sloshing as the car accelerates and decelerates (something to think about if you roll your own)

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----

formatting link
The #1 Newsgroup Service in the World! 120,000+ Newsgroups

----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Reply to
default

Wouldn't an inverting unity gain amplifier and some level shifters be the solution? Since the gauge has a higher impedence than the sender, you should not need power amplification, only level shifting and inversion, and the power demands of the gauge seem okay for many common op amps.

Reply to
Don Stauffer

You can't do it by rewiring the thing, because it responds differently to the input. You can put a test current through it, measure the resulting voltage, and use that to create a resistance using various analog schemes, or perhaps a mechanical pot.

If you are handy with a micro controller, you can use an on-board A/D to measure the output of your sender (by putting a known current through it and measuring the voltage), and use that to control a 2 wire digital pot. The nice thing about this solution is that you can easily calibrate the result using a software lookup table, allowing for a more accurate display. This is a 3 hour project if you have a PIC, programmer, digital pot, and linear regulator handy...

Reply to
Bob Monsen

How much is it worth to you? Get someone to take a $3.00 pic micro, Read the resistance of the sensor, then translate it linearly, and output to a 8 bit Digital Pot for the Dash Gauge with the new resistance. $10.00 in parts, a few lines of code, and hopefully a buddy with a programmer. This is my easy way out however others may have a easier way using components only.

Richard

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----

formatting link
The #1 Newsgroup Service in the World! 120,000+ Newsgroups

----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Reply to
Richard

This sounds by far the most flexible and accessible solution so far, especially given that my main area of expertise is in computers. Other solutions involving analog components, inverters etc all sound too complex for me.

Are there any PICs around with integrated voltage regulator? I'll have to feed it off 12-15V. Plus, I couldn't find any digital pots in 500Ohm range - they all seem to start at 10K. Any ideas here?

Thx! Peter

Reply to
Peter

If I understand correctly this will only change resistance range. What I really need is for resistance to gauge to increase while sender's resistance decreases. Or am I getting something wrong here???

THX! Peter

Reply to
Peter

I had a car that wasn't like that. The fuel display was a crossed coil device with a fixed resistance on one axis and the sender on the other axis. (fairly standard in older vehicles)

it responded to the sloshing of fuel in the tank and when the fuel level was low you could judge it down to the last cupful by the amount decelleration needed tp get a twitch out of the needle.

i didn't note a damping capacitor on the instrument cluster or elsewhere on the schematic.

Bye. Jasen

Reply to
Jasen Betts

you could use a cheaper micro and build tje dac in software using a by measuring the discharge time of a capacitor.

the output could also be done in software using a PWM output to drive a power fet and a smoothing cap.

i'm not sure wether the reduction in pin count and expense of partsa is worth the extra software complexity or not

Bye. Jasen

Reply to
Jasen Betts

Your problem begs the question as to why the fuel level sender does not match the requirements of your fuel gauge. One would think that if all the components were already in a vehicle (or whatever they are in), the gauge would respond as dictated by the fuel level sender.

That aside, there are apparently 3 main types of fuel level senders in common use. Read the FAQ from Stewart Warner here

formatting link

From your description, you have the GM type sender whereas you really need the Stewart Warner type.

Reply to
Ross Herbert

This is dual-fuel vehicle that runs both on regular gasoline and liquified propane gas. LPG sender is the one I'm having problems with... it's not user-replaceable and is inside pressurized container.

I'm starting to think that path of least resistance will be to add separate

0-90 Ohm fuel gauge...

Peter

Reply to
Peter

Makes no sense - a magnet could replace one coil cheaper unless it is a sine-cosine resolver - then both coils are driven with signals (not a fixed source)

Maybe there is no "standard" fuel system.

My old MG midget showed the slosh - want the fuel remaining? drive on a level road at a constant speed or park on the level.

Every light aircraft I've driven has the same system. (shows the slosh)

But the old ("big3") cars used a system of indicator and voltage regulator that depended on a bimetallic heater to regulate the voltage and indicate the level (there was a heater in the gauge). Shut the car off and the gauge drifts down to zero. Those were common from the

30's to 80's.

The newer american cars have a heavily damped indicator, that maintains the indication with the engine off (at least from what I've seen since about 1980) That system acts something like a peak level indicator.

Crossed coils doesn't make sense, unless you're talking about an indicator that can go 360 degrees? or more than ~270? Sine-Cosine resolver.

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----

formatting link
The #1 Newsgroup Service in the World! 120,000+ Newsgroups

----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Reply to
default

As to the regulator, just buy an 7805. Don't forget the bypass caps (look at the datasheet for the appropriate values).

Do you know how the display works? How much current does is pass in normal operation? It might be possible to instead use a negative impedance converter (although you would also need a negative voltage source for this.) Currents more than a few mA will probably not be possible with a circuit like this, though. Also, I don't now how to scale the output of one of those, although I'm sure somebody like bloggs can figure it out.

--
Regards,
 Bob Monsen
Reply to
Bob Monsen

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.