Driving backwards with a boat (Rover) in Manual mode

stefanmessmer · January 6, 2022, 8:29pm

Hello
We have built a boat with two motors configured in differential mode. When using the boat in manual mode we should have the possibility to drive backwards. Up to now the motors are stopped at the minimum throttle position (I think this has been configured by calibrating the ESC’s). If we turn the boat with throttle = 0, one motor goes forward and the other goes backward. The ESC’s are therefore correctly configured to operate the engines in both directions.

What is the most elegant way to solve this problem?

Many thanks and best regards
Stefan

stefanmessmer · January 12, 2022, 7:59pm

Hello
We have found an elegant way to solve the problem: Setting RC3_TRIM to 1500 us allows to change the direction of the motors and driving backwards in manual mode. Don’t forget to move the throttle stick in the middle position before arming .

Best regards
Stefan

heilmao · February 7, 2023, 1:13pm

This worked for me, thanks!

junwoo0914 · March 4, 2023, 9:43pm

Are you building a boat?

heilmao · May 1, 2023, 5:19pm

Hi @junwoo0914, I am building a car! Can confirm that this approach works with other vehicles besides boats. Thinking about it, this makes sense as its all just motor control with ESC right?!

Perhaps this clarifies things to others landing here too…

junwoo0914 · May 2, 2023, 9:05am

Yes, Boats and Rovers essentially have same conceptual design (reversible motors & steering).

On the side note, the solution presented here is probably coming from the fact that in QGC there’s a hotfix that has always resetted the throttle channel’s TRIM value (e.g. commonly RC3_TRIM) to the minimum value (e.g. 1002), and not the mid-range value (1500).

This is further explained in this comment:

Therefore the fix to revert the hotfix would be needed: Hotfix for sprung-loaded throttles and RC calibration: Enforce correc… · mavlink/qgroundcontrol@0577af2 · GitHub. For that the fix PR has been sitting a bit here:

github.com/mavlink/qgroundcontrol

Remove RC Throttle trim calibration value getting hard-set to Minimum value

mavlink:master ← junwoo091400:pr_remove_rc_throttle_trim_hotfix

opened 09:30AM - 19 Nov 22 UTC

junwoo091400

+0 -7

## Descrption A hotfix [commit](https://github.com/mavlink/qgroundcontrol/commi…t/0577af2e944a0f53919aeb1367d580f744004b2c) from 7 years ago was forcing RC transmitters to always have a trim value set to 'Min', effectively only allowing the [0, 1000] range. ~~While for legacy reasons & to bypass cases of spring-loaded transmitters, this may be ok, it is certainly not true regarding reversible thrust vehicles (e.g. Boat / Rovers), and it is confusing why only the throttle would be in the range [0, 1], and not [-1 ,1], as defined in the [`MANUAL_CONTROL`](https://mavlink.io/en/messages/common.html#MANUAL_CONTROL) MAVLink message definition.~~ ## Changes Removed this hotfix, and allows setting the trim values automatically in the state machine (need to check logic once again) ## Context The discussion that discovered this flaw is here: https://github.com/PX4/PX4-Autopilot/pull/15949#issuecomment-1318728541 ~~We need to also clearly define the point about the definition of `MANUAL_CONTROL`'s throttle (`z`) definition. As it doesn't really make sense to have throttle vector's orientation interfere with the RC transmitter's setpoint values (it should be agnostic to vehicle's behavior to be basic).~~ Also, PX4 will be able to handle this change using this logic (although some legacy parameter translations / user awarenesses would be needed): https://github.com/PX4/PX4-Autopilot/pull/15949 And we need to make sure for ArduPilot this will work out as well. Judging from [the way Steer / Throttle Manual override isn't differentiated](https://github.com/ArduPilot/ardupilot/blob/b2a38c0c6033a56bd4e7f7bffe65b70381d6ea76/Rover/GCS_Mavlink.cpp#L814-L815) in Rover code, I guess ArduPilot already uses full range (-1000 ~ +1000) definition for the internal RC throttle data, @peterbarker. Did you not have problem with QGC, regarding this point? ## TODOs - [ ] I am still unclear how the rcTrim value would then get set for throttle, need to check the state machine logic to be sure.

MaEtUgR · May 5, 2023, 9:36am

Thanks for the report and posting how you solved it!

Setting RC3_TRIM to 1500 us

In newer versions of PX4 I do this automatically:

github.com

PX4/PX4-Autopilot/blob/40324b03f4acc8eb49423b4e12b585a0cabfd713/src/modules/rc_update/rc_update.cpp#L205-L218


      
          	// Center throttle trim when it's set to the minimum to correct for hardcoded QGC RC calibration
          	// See https://github.com/mavlink/qgroundcontrol/commit/0577af2e944a0f53919aeb1367d580f744004b2c
          	const int8_t throttle_channel = _rc.function[rc_channels_s::FUNCTION_THROTTLE];
          
          	if (throttle_channel >= 0 && throttle_channel < RC_MAX_CHAN_COUNT) {
          		const uint16_t throttle_min = _parameters.min[throttle_channel];
          		const uint16_t throttle_trim = _parameters.trim[throttle_channel];
          		const uint16_t throttle_max = _parameters.max[throttle_channel];
          
          		if (throttle_min == throttle_trim) {
          			const uint16_t new_throttle_trim = (throttle_min + throttle_max) / 2;
          			_parameters.trim[throttle_channel] = new_throttle_trim;
          		}
          	}

It’s working around the “hotfix”/hack in QGC which @junwoo0914 mentioned and we need to get rid of.

Related to that I just recently found that QGC has an extra case for vehicles like rovers and boats where it just changes the range of joystick values:

github.com

mavlink/qgroundcontrol/blob/7de48208c312392a916ddf8284db64e36a2f896a/src/Joystick/Joystick.cc#L668-L670


      
          if (!_activeVehicle->supportsNegativeThrust() || !_negativeThrust) {
              throttle = std::max(0.0f, throttle);
          }

github.com

mavlink/qgroundcontrol/blob/7de48208c312392a916ddf8284db64e36a2f896a/src/FirmwarePlugin/PX4/PX4FirmwarePlugin.cc#L707


      
              Q_UNUSED(vehicle);
              return QStringLiteral("https://api.github.com/repos/PX4/Firmware/releases");
          }
          
          QString PX4FirmwarePlugin::_versionRegex() {
              return QStringLiteral("v([0-9,\\.]*) Stable");
          }
          
          bool PX4FirmwarePlugin::supportsNegativeThrust(Vehicle* vehicle)
          {
              return ((vehicle->vehicleType() == MAV_TYPE_GROUND_ROVER) || (vehicle->vehicleType() == MAV_TYPE_SUBMARINE));
          }
          
          QString PX4FirmwarePlugin::getHobbsMeter(Vehicle* vehicle)
          {
              static const char* HOOBS_HI = "LND_FLIGHT_T_HI";
              static const char* HOOBS_LO = "LND_FLIGHT_T_LO";
              uint64_t hobbsTimeSeconds = 0;
          
              if (vehicle->parameterManager()->parameterExists(FactSystem::defaultComponentId, HOOBS_HI) &&
                      vehicle->parameterManager()->parameterExists(FactSystem::defaultComponentId, HOOBS_LO)) {

This is disturbing in my eyes since the ground station should not need to know how it has to scale the joystick before being sent to the autopilot depending on vehicle details. It should rather send clearly defined joystick input and the autopilot which has the context of vehicle, mode, configuration can then interpret it the right way.

The problem is the seamless transition away from such mistakes from the past.

Tink · June 5, 2023, 7:19am

Hi MaEtUgR, I’m encounting the same question as described above.
I noticed that in the firmare v1.14.0, rc3_trim is already defined 1500 PARAM_DEFINE_FLOAT(RC3_TRIM, 1500);
But with this version I still cannot drive the motor reversely with the RC stick. Do I also need to set the parameter in QGC UI?
Thanks a lot.

junwoo0914 · June 13, 2023, 11:57am

Does your RC transmitter by any chance have reversed throttle channel?

github.com/PX4/PX4-Autopilot

rc_update: throttle trim centering fix for reverse channel

PX4:main ← PX4:maetugr/rc-throttle-trim-reverse-fix

opened 12:36PM - 05 Jun 23 UTC

MaEtUgR

+6 -2

### Solved Problem When integrating an airframe I found that if the throttle ch…annel is reversed the logic that corrects for QGC calibrating the throttle to half the range (see https://github.com/PX4/PX4-Autopilot/pull/15949#issuecomment-1318728541) ### Solution - Take care of the case where the channel is reversed because then QGC sets the trim parameter equal to the maximum channel value to achieve the same half range that was in use before #15949. ### Changelog Entry For release notes: ``` Bugfix: Get full throttle range for QGC RC calibration with reverse throttle channel ``` ### Alternatives This can be removed again once the calibration in QGC is revised and available in a stable release for long enough. ### Test coverage - I tested this on the real vehicle with SBUS and a reversed throttle. - I also double-checked that the unreversed normal case still works as expected on the same setup by changing the parameters to a non-reversed throttle case.

This PR recently fixed this exact problem occuring for reversed channel transmitters, and to try this out you can upload the latest main branch firmware in QGC and try it out again.

Could you be able to do that?

Tink · June 13, 2023, 11:58pm

hi junwoo
I find that the RC_TRIM parameter in PX4 code is 1500
However in QGC this parameter was set to 1002 after RC calib.
I chaged it back in QGC and it works well now.
Thanks~