Bar30 fail– depth 0 in QGC and -2900 hPa value in autopilot sensors

We are away on fieldwork with our BlueROV, so any speedy help with this would be greatly appreciated.

Our set-up is a BlueROV heavy (original version), that has been upgraded with a RPi4, Navigator flight controller and BlueOS.

All was working well, until its last dive with depth showing 0 though out. Checking the vehicle set-up in BlueOS, we can still see the Bar30 sensor, but its giving values way out. Approximately -2900 hPa at ground level.

We have tried applying a small amount of air pressure to the sensor using the leak check pump and the raw values decrease correspondingly, suggesting that the sensor still functions, but is giving strange values. We have tried to calibrate it, but it does not improve. Does anyone have any ideas? Perhaps there is a variable we could adjust to apply an offset?

Please see the screenshot below from BlueOS.

Hi @JohnH -
Was the sensor used in water for a continuous period longer than 24 hours? It does seem as if your sensor has been damaged. Please reach out via our support form with your order # to continue troubleshooting and discuss a replacement.

Hi Tony,
Thanks for your reply. The ROV has been freshwater washed then towel dried after each days field work. Will see about replacing it when we get back to base.

Interestingly, with the ROV disassembled, we have found the sensor randomly fluctuates between sensible values and not. I’m wondering if it could be some kind of power supply/electrical noise issue.

When you have seen these sensors fail in the past, have you ever seen this kind of fluctuations? I’m still holding out for a fix in the field!

Many thanks.

Hey John
Also I’m the field on my end too. We had a similar issue a few months ago and it turned out to be electrical noise from some modifications we made.
We solved it by adding a ferrite core to the i2c cabling.
See our forum post Here

Good luck,
Mike