OP2 pypilot driver_timeout - Printable Version

+- OpenMarine (
+-- Forum: Pypilot (
+--- Forum: General discussion (
+--- Thread: OP2 pypilot driver_timeout (/showthread.php?tid=2669)

OP2 pypilot driver_timeout - AlanH - 2020-06-21

Sean et al,

Thank you for the excellent OP2 upgrade! I'm excited for the Signal-K integration.

So, some problems getting pypilot to play nicely ... Starting with a NOOBs Headless image, I installed newest pypilot as of last week (2.0.8beta I think?). IMU data is good, gps is good, OpenCPN plugin overlay is working.  But when I select AP it throws DRIVER_TIMEOUT.  No servo data on Statistics or Scope.  (see attachments)

Reverting to 1.2, the motor works normally so it's not hardware.

What am I missing?


RE: OP2 pypilot driver_timeout - AlanH - 2020-06-23

       # if not moving or faulted, disengauge
       if not speed or self.fault():
            #print('timeout', t - self.command_timeout)
I find myself wondering ... since I have no position sensor, is it choking when it tries to infer rudder speed?


RE: OP2 pypilot driver_timeout - seandepagnier - 2020-06-24

DRIVER_TIMEOUT is really just a warning. It indicates that the motor controller is driving the motor but no current is measured in the motor. It usually occurs if the motor is disconnected but the controller is working.

I would check the connections and make sure manual control is working. If not, try backing up and then deleting pypilot.conf, maybe there is a setting in there that is preventing the motor from moving correctly.

RE: OP2 pypilot driver_timeout - AlanH - 2020-07-03

Manual control works. Motor control works in pypilot 1.2.

Ok, I think I found the problem: Bad IMU.

pypilot detects the IMU and receives data at first, but then zeros out.  See screen shots.  I think the crashing problem before was when it failed to detect the IMU.

Unfortunately, I'm in the middle of a 10-day trip. Is there any way to force navigation using only gps data?


RE: OP2 pypilot driver_timeout - seandepagnier - 2020-07-04

Unfortunately the IMU is required. Most of the steering correction is from the gyros... really the other sensors just correct the small errors in the gyros. Without gyros, the turn rate would have to be emulated and it would never work very well

The next version will support more pilots. For example the wind pilot does not use the compass sensors, it uses only wind sensors and gyros so compass errors/distortions wouldn't affect it. It would be possible to make a pilot that only uses gps and wind or some other combination but I don't expect good performance.

I'm concerned about these bad sensors. Maybe the pi you are using gets very hot compared to a pi zero and this leads to sensor failures? It would be good to determine the reason, if the sensors are really broken, and if they are damaged forever, what the cause is?

You said it worked with the old openplotter 1.2?? this doesn't make sensor for the sensors but maybe the hardware serial port wasn't configured in op2 and this is why the controller didn't work?