Need help with solution for digital resistor

[lovemb]

Hi everyone,

basically I'm creating a device to control various automotive dashboards (in the future, more than one at the same time). On newer dashboards most stuff is simple because info is sent by CAN, however on old units many sensors are variable resistors (mainly temperature sensors and tank fuel level sensors).

I need a way to simulate these sensors and control their value with a microcontroller.

I'm aware of digital potentiometers but there are no pots that can emulate in the resistance I need.

Usually 5 ohms to 200 ohms.

Here are some examples of fuel level sensors:
240-33 Ohms - Standard Aftermarket -  T-SW-**
10-180 Ohms - VDO gauges - T-VDO-**
0-30 Ohms - GM up to 1964 -  T-GME-**
0-90 Ohms - GM 1965 and up - T-GML-**
73-10 Ohms - Early Ford & Mopar - T-ORG-**

Example of the circuit and component I'm emulating:





Usually one side of the component is connected to ground, but not always.

There are 1k digital pots, but at 256 steps and a wiper resistance of 50 ohms is not usable.

Motorized potentiometers could be an option, but I can seem to find 500ohm or 1k ones.

Also it can't be too expensive because I need to create many of these controllers for same project.


Any ideas?

[Someone]

Also it can't be too expensive because I need to create many of these controllers for same project.

Wait until you see the currents some of those sensors pass through them!

[Benta]

I'd suggest mini/micro servos for RC models (cheap and plentiful) mechanically coupled to loudspeaker rheostats (2...5 W, freely available in 100 ohms), also cheap.
Controlling the servos is super-easy.
.

[moffy]

If you PWM a low value resistor the effective value = R*(Ton+Toff)/Ton. You would need to smooth the current somehow, not sure at present if it is practical.

P.S. I don't think it would work in this case.

[Swainster]

Perhaps a programmable current sink would do the job

[Someone]

If you PWM a low value resistor the effective value = R*(Ton+Toff)/Ton. You would need to smooth the current somehow, not sure at present if it is practical.

P.S. I don't think it would work in this case.

It does....
https://www.eevblog.com/forum/projects/creating-controlled-resistance-to-ground/?all

[MasterT]

Yes, programmable current source. Same as electronic load
https://www.electronics-lab.com/400w-1kw-peak-100a-electronic-load-using-linear-mosfets/

Since current is low, one opamp like lm6500 or TCA0372 would be sufficient. Probably uCPU & DAC could be powered from the same line - 2 wire connection.

[ledtester]

Check out this recent thread:

https://www.eevblog.com/forum/beginners/digital-adjustable-resistor/

I think the idea of switching in/out resistors in series with relays or MOSFETs is the most practical solution, for instance:

https://www.eevblog.com/forum/beginners/digital-adjustable-resistor/msg4439026/#msg4439026

Another idea which uses resistors in parallel switched to ground is mentioned in this post of mine:

https://www.eevblog.com/forum/beginners/digital-adjustable-resistor/msg4451482/#msg4451482

It references the op-amp circuit in reply #52.

Also, a device like this might be of interest to you:

https://www.eevblog.com/forum/crowd-funded-projects/high-accuracy-programmable-cute-resistance-substitution-box/

Controllable from a computer makes it suitable for automated testing.

Check the wattage of the resistors used -- I think they are 1W.

[Doctorandus_P]

Look at the bigger picture.
If the original sensor resistor is connected to GND, then it probably pulls a current from the other side. That means that your sensor emulator does not have to be a variable resistor itself, but it only has to sink a current to GND, and that can be done with an adjustable current sink, which is easy to make with an opamp, a transistor and a small value shunt resistor.

Edit:
I see now that a programmable current sink has already been mentioned twice, so consider this the 3rd time

[inse]

If you want the meters to show specific values, it will be an additional challenge.
The thermistors do not have a linear characteristic and also the fuel level sensors may be a little more complicated than the wiring diagram shows.

[David Hess]

A programmable voltage or current source will work, which could mean buffering the output of a digital potentiometer or digital-to-analog converter.

Alternatively if you want a true variable low resistance, then a 3-wire digital potentiometer can operate as a multiplying DAC and that can be buffered to make a variable low output resistance by using the output voltage as the input to the digital potentiometer which then controls a current.