Wind Vane's potentiometer values change depending on power source

I have a Davis Anemometer that I’d like to connect to my arduino uno weather station.

I used the guide stated here: Davis Anemometer Arduino Hookup | Code (Part 1 of 3)

When I tested the wind vane at the bench, the values that came from the pot seemed in line with what the guide stated.
However, on my weather station I have a power bus that feeds all of the components, as well the the MCU.
It can be powered from the USB on the arduino, or the solar panel/battery or whatever.

When I power the Wind Vane from the bus, the pot values are way off, and seem to taper from 0 - 1024.

I imagine that it has something to do with a different resistance in the bus as opposed to right from the arduino 5V pin.

How do I proceed?

Edit: After some testing. Even having the bus ground plugged into the arduino, and running the wind vane directly from the Arduino 5V/Gnd, tends to send the wind vane’s values way out of wack.

Edit: It’s the I2C bus that’s connected to the power bus. Isolating the I2C bus, causes the wind vane to display the correct values.

Edit: Plugging any I2C device into the I2C bus,(which takes it’s 5v/GND from the power bus mentioned above) and then connecting the SCLK and SDA lines, causes the issue.

As soon as I unplug the data and clock lines, the wind vanes values switch back to normal.

Edit: Do I need to maybe connect a lead from the 5V power bus to the AREF?

Fritzing attached

A4 and A5 are used for I2C.

The example uses analogue input A4 for wind direction.
Did you change that to a different input.

I am using the dedicated SCLK and SDA pins on the arduino R3.
But this could certainly be a clue.

I was not aware of the fact that A4 and A5 had some relation to I2C.
When it stops raining, I'll switch the pins and give it a shot.

You were right Leo,
Moving the wind sensor to A0 instead of A4 or A5 worked.

I guess I never figured that the SDA and SCK pins were wired in parellel to the Analogs.

Thanks much.