Skid Steer Rover going in reverse in Auto / Guided

ArduRover Rover 4.7
1

Has anyone seen this issue.

I have a skid steer wheeled rover that drives normally in manual mode. When I switch to auto for waypoint mission or guided it drives backwards instead of forwards. The m10 holybro gps has good connection to satellites and a 1.0 HDOP. The orientation of the GPS is correct on the HUD in relationto the vehicle. It also tracks properly with yaw adjustments. For reference I am using a roboclaw to drive 4 motors. Left side motors spin the same direction, and right-side motors are wired to spin the same direction.

Update - Here are some params and the values
SERVO1_REVERSED = 0
SERVO3_REVERSED = 0
RC3_REVERSED = 0

I am out of ideas. Thanks in advance for any input.

2

It sounds link there’s a reversal somewhere in your system. Try running in ACRO mode and supplying a logfile

3

Thanks Stephen. I will work on that when I get in the office. Ill upload in a bit.

4

Moved to 4.7 (and removed duplicate topic).

5

Stephen -

I did a test in acro mode. as soon as I switched from manual to acro it went full speed in reverse.

Link to log File:

6

Hi @badgersix,

This is a common issue with new setups. I guess the important step of testing the motors order was skipped?

Could you try that and make sure the left and right motors are spinning in the correct direction?

7

As far as I can see from your log there are no changes on RC Throttle channel. So produce a log showing some action in manual and some action in an auto mode.

8

Motor test in Mission Planner was successful

Motor c test produced left wheels going forward
Motor D test produced Rignt wheels going forward

9

It’s avoidance behavior. You have a prox sensor enabled with a margin of 2m and reverse speed of 0.75 m/s. The sensor shows an obstacle at 0° azimuth about 0.5m away for the duration of the log, and the Rover is trying to reverse away from it.

Disable avoidance behavior for now (AVOID_ENABLE,0). Accomplish the tuning procedures such that the vehicle is well behaved in all modes, and then revisit avoidance later.

10

Thank you Yuri!
I will test this afternoon. That actually makes the most sense. I need t get better at interpreting logs.

11

That one was very insidious. The keys to noticing something beyond basic analysis were:

  • No servo or RC channel reversals.
  • All servo and RC trims set to sane/expected values.
  • No RC3 (throttle) command while ACRO mode demanded saturated reverse throttle.
  • Correlation between desired speed and avoidance reversal speed.

Catching something like this is actually pretty difficult, even with a trained eye. I stumbled into it.

12

I appreciate your help. I owe you a coffee.

13

I changed the setting you posted and that fixed the issue. Much Thanks!