Certainly understand the concept of positive pressure keeping the water out. But was just thinking of extra safety step in case there is a problem with pump or intake valve.
In regards to the baro error I’m surprised the parameter even exists in the rover version. In the copter version there is a parameter “Ek2_ALT_Source” you can find in MP or using QGroundcontrol - whichever you prefer. Whilst that setting can’t be turned off you could try changing it to GPS or any other sensor just to see if that stops the message. Another possibility is in the failsafe settings it may be activated by standard and you just need to disable it.
There is also a " BRD_ALT_CONFIG" setting but apparently the options there depend on the control board being used.
In regards to using the kakute baro to control pressure - I’ve got no idea. Some other control boards can communicate with the arduino board you’ve mentioned and with a small script the data could be read and the pump triggered when required. On that board I have no idea. - Sorry.
I personally am much more inclined to search for a simple pressure switch (barometric switch) in order to control the pump. Much more reliable than a mini computer, software and everything else that would go with that.
…half an hour later:
Have given your project more thought…
The reasons for having a barometric sensor reading in ardurover are most likely:
a) As most control boards have an inbuilt sensor for air pressure, why not use it?
b) If you decide to sent a rover up a steep hill you may want to know how high it is / has been.
this brings me to the problem with the error message you’re getting. I suspect what is happening:
a) the pressure you create with your pump is higher than what the sensor is designed to handle or the values received are higher than expected by the board.
b) The pump you’ve shown in the photo appears to be a small diaphragm or piston air pump. Those kind of pumps don’t create a steady flow of air like a fan would, but rather provide airflow in pulses. If the barometric sensor is sensible enough it will pick up on those airflow pulses and hence creates a error state as such a thing would never happen in an aircraft.
So, if my assumption is correct you can’t use that sensor for for controlling the pump.