I have an issue sensing the throttle position sensor voltage if the sensor is already being used by the vehicle's own ECM. Namely, the vehicle ECM stops responding to the throttle input after a period of time from 20 seconds to 5 minutes later.

When I probe with my multimeter (positive lead on the TPS #1 Signal and negative lead on the TPS #1 Return), the voltage at the pedal goes from 0.5V to 4.5V. This doesn't affect the vehicle's ECM at all.

When I use my ATMEGA32u4, connecting the TPS #1 Signal to PF0 and the TPS #1 Return to my AVR's GND network, the vehicle ECM complains about not having a throttle connected.

What should I consider adding between the TPS #1 Signal and the AVR to make it 'disappear' in the circuit and not affect the vehicle ECM? Or is this a function of the internal pull-up resistor?


1 Answer 1


Or is this a function of the internal pull-up resistor?

More than likely it's the internal pull-up resistor you have on the AVR. If you couldn't remove that pull-up (for whatever reason) I'd consider using an op-amp as a unity gain buffer. Choose an op-amp that can run from a 5V supply, and has rail-to-rail inputs and outputs for least problems.

enter image description here

Don't forget to wire 5V power and ground to the op-amp. Maybe an AD8605 could be a good start - it's over the top for what you want but will give you an idea.

  • \$\begingroup\$ I ended up with a TI OPA2348AIDCNT to test adding a unity gain buffer for both sensors that I'm reading. \$\endgroup\$
    – joefarmer
    Commented Jun 26, 2014 at 19:24
  • \$\begingroup\$ Was the op-amp OK for you? \$\endgroup\$
    – Andy aka
    Commented Jun 26, 2014 at 19:38
  • \$\begingroup\$ I had to order it, so I'll have an answer once I test it in a few days. I was not able to disable the pullup when using the ADC, even though the documentation indicates that it would be turned off. Also, it's nice to have a buffer between the MCU and the automotive electrical environment. \$\endgroup\$
    – joefarmer
    Commented Jun 27, 2014 at 17:22
  • \$\begingroup\$ The TI op amp works perfectly in the final design. Thank you again. \$\endgroup\$
    – joefarmer
    Commented Oct 3, 2014 at 16:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.