Pixhawk V3 AUX Pullup or not

I am trying to add a flow sensor to the Pixhawk GPIO as described in Fuel Flow and Level Sensors — Rover documentation but face the challenge of getting the right Voltage levels. Which i guess is related to a Pullup-Pulldown issue.

My sensor provides a nice PWM in a “dry” setup with a 1,5k pullup 5v power supply.
However, when plugging into the Pixhawk V3.0 the pullup seems to fight an internal pulldown. The voltage drops to about 0.9V. It still provides a good pwm but with the wrong levels.

After some searching in the docu I tried 2 things so far: Base on Buttons — Rover documentation I tried removing the pullup completely since it states that

These pins will have an internal pullup assigned to them automatically,

that did not work. It was a bit like a very noisy PWM on 5V level.

based on Fuel Flow and Level Sensors — Rover documentation I tried a pullup to 3.3V with a 10K even though that does not comply with the specs of my flow sensor.

Any GPIO capable pin on the autopilot can be used to connect to the sensor’s output. If it is an open-collector output, a 10Kohm external resistor pull-up to 3.3V will be required

This also did not give me the desired result just a very low level PWM at around 0,3V

I have no idea about the inner schematics of the Pixhawk, maybe someone in this forum has a good idea, thanks!!

Hi @monkfood Welcome!
Well since you are using Ardupilot you should ask this on their forums.
No One here is that familiar with Ardupilot.

Hello, thanks for your suggestion. But since this is not an issue of the software but the Hardware, it should be correct here.

But today I got the solution. Apparently I was terribly wrong in the assumption that pixhawk is a shared design like an ARM chip. Meaning that, independent of the producer, Pixhawk of a specific version from equals Pixhawk of the same version from another producer. I now know that the one I was able to aquire does not have a level shifter and pull up. So next time I know that to check for before buying.

The I/O pull ups and down can be under firmware control at least on the STM chip , so it could be software controlled as well. Which is why I made that suggestion - it could be different.