Increase resistance in a circuit by a percentage

[chasarms]

First of all, I am newb here. So, if I am violating any sort of forum protocols, I apologize in advance. As well, I am very grateful for any insight that anyone might offer.

My question is this;

Is there any sort of circuit out there, (preferably passive but I could supply a 12-14V+ DC if needed) that would allow the resistance on a wire to increase by a percentage?

My problem is that I have a sensor in a 12V+ DC automotive system that is "underoutputting" for the calibrations of the visible gauge. That is, the resistance on the sending unit is not as high as needed for any given position.

For example if the gauge is expecting a value of 50 ohms on the wire in order to read the mid point of say 100 degrees, the sending unit is only providing 40 ohms when the condition of 100 degrees exists. So, the gauge is not accurate.

Before you suggest simply adding a resistor in series, note that the sending unit doesn't behave in a linear way. I have run some tests, and ideally, i would like to increase the resistance on the wire by 15-20% at any given position of the sending unit.

Other restraints are that I cannot access the sending unit itself. I can access the gauge. It is a standard automotive style with a 12v+, a ground and the wire from the sending unit.

Lastly, note that there is no expectation that replacing the sending unit would result in better performance. It is a known issue with the system. I just want to improve the reliability of the information from the gauge.

Also, I am fairly dexterous, can solder well enough to get electrons to flow and not burn down anything, and I am smart enough to know that my chances of help here are far greater than at any automotive forum, but I don't have a great deal of experience reading circuit diagrams, so dumb it down as much as you can. Many, many thanks in advance.

 

[JoeJester]

For example if the gauge is expecting a value of 50 ohms on the wire in order to read the mid point of say 100 degrees, the sending unit is only providing 40 ohms when the condition of 100 degrees exists. So, the gauge is not accurate.

What are the tolerance specifications on the sensor (sending unit).? Right now your talking about +/- 20%

 

[Harald Kapp]

I can't envision a passive way that at the same time matches the transfer curve of the sensor. A thermistor in series with the sensor and in thermal contact with the sensor could track the sensor's temperature and add a corrective value to the sensor's output. But It will be very difficult if not almost impossible to match the resistance-vs-temperature characteristic of the thermistor to that of the sensor.

And alternative could be to disconnect the sensor from tha gauge, measure the sensor's resistance e.g. by a microcontroller, perform any corrections on the sensor signal and output the corrected signal as a simulated resistor (e.g. a voltage controlled resistor - Google) to the original gauge. But that is not as trivial as it sounds.

 

[BobK]

An analog way might look like this:

1. Run a fixed current through the sensor to convert to a voltage.
2. Amplify this voltage via an op amp.
3. Use this voltage to control a current source connected to the meter.

I have no idea whether this is workable or not

Bob