I'm starting this topic because there was a topic in an analogous device in the main forum. The OP needed a digital potentiometer over the range 50-600 Ohms and 12V. Turned out a rheostat was what they were actually needing but the problem was that no semiconductor digital device was known that could handle the power.
The device it is to replace somehow varied the resistance over temperature to drive a gauge from 95ºC to 260ºC. OP refused to describe the original sending unit.
As an old motor head whose family have been in the business for over a century I got really curious about what this device must be. It is obviously air-cooled if it's an ICE based on the temperature range. Probably a vintage 2-stroke of some sort. I haven't been able to satisfy my curiosity via a web search.
So, how do it work? Does anyone remember those thermostatically controlled dampers that were activated by an expanding bellows? I can envision one of those driving the kind of rheostat in an automotive fuel level sensor. That would handle the power okay.
From an old post from the OP, this might be from a vintage British bike if that provides a clue.
But the gauge at the end will measure the current, right? Or the voltage across some shunt resistor in series. Or could it be another thing?
Then, beside the PWM solution, wouldn't it be possible to drive a mosfet or a FET, (or a LM317), with an opamp? as a constant current device controlled by the opamp and a MCU DAC pin.
It would look like a variable resistor and would be able to handle the power, I think.
I'm trying to figure out how it was done 75 years ago when there were no FETs or Arduinos.
A sending unit that was mounted in contact with the engine cylinder head that looked like a 50-600 ohm resistor over a temperature range of 95ºC - 260ºC. The resistor had to safely dissipate several watts.
The sensor does not exist. Any way that should be a NTC Thermistor of 50 to 600 ohms at the right temperatures.
I understand from the detailed and not very irritated answers, that he already has a way to measure the temperature and just need to trick the gauge in some existing panel. But no idea of what could be the original sensor.
That is what the OP from the other topic wants to do.
The solution they have settled on definitely is influenced by their belief that they have to switch actual resistors in and out of circuit. They wanted compact but have decided that 12 or more MOSFETs and a multiplexer are more compact than PWM and a single MOSFET. That's fine, it's their project.
After doing some more searching I've come to the conclusion that a high power low resistance thermistor is not a unicorn after all and that must be the device that Triumph, Norton, etc. used.
And doesn't necessarily have to dissipate a lot of power depending on the gauge it's driving. As I learn more about NTC thermistors in that range it looks like power dissipation is not the problem I thought it was.
Understood. The 100 ohm "resistor" in the sending unit that has to dissipate ~1.5 W in a high ambient temperature is the part that I didn't realize was so common. These days, 10k thermistors are more the norm so I don't immediately think of high power.
The gauge is 100 ohms, the sensor is 100 ohms, they are in series with 12V across them thats not 1.5W.
There was no standard temp sensor the resistances varied from make model and year.
Nope. Just thinking OP might not know what is really needed.
I had a '68 Volvo which used a wirewound resistor as the sending unit from the gas tank and drove a resistive 'thermostat' kind of thing in the dash. As I remember it was a primitive sort of PWM which sent current to the gauge on the dash. The meter response was so slow it just looked like analog.
Definitely. Neither a skilled mechanic nor knowledgeable about things electrical based on other posts from same.
That was sort of the basis of the kind of sensor I was envisioning for temperature. I have since found that a thermistor with the proper credentials must exist.
