I’m working on a camera/robotics project with a Permobil wheelchair as base.
Taranis TXRXSBus -> Pixhawk 2.1/Cube -> Roboclaw Controllers via PWM
Both input channels are coming from left gimbal, setup as left right axis for Ch1, vertical axis for Ch2
Routing the RC inputs directly to “RC Pass Through”, instead of Throttle Left 73 and Right 74 gives me PWM output to my controller.
But configuring as described, my Ch1 and Ch2 inputs just go into a black hole, and I can not get any PWM output.I suspect there is something I haven’t picked up in the docs about how ArduRover works with arming, etc. Any parameter I’ve found that looks like it would inhibit motors turning as a failsafe, I’ve disabled. Ive been trying without a GPS unit, that comes tomorrow although I doubt it’s my issue.
BTW, I believe the 4 RC_Map params are set correctly for my CH1 & 2 input scheme.
What params should I be looking at? I feel like there might be a lone one somewhere hanging me up.
Well, I got it working. A search indicated that somebody suspected skid steer is implemented in a sort of hard coded way, where the function is only initiated when Serv1 is set to 73, 3 is set to 74 (just as instructed), but at that point if the rest isn’t just as documented (RC1 and 3, etc etc), the skid steer controller won’t instantiate and what input you do have goes into a black hole.
Great looking rover and great that you got it working.
Sorry for not seeing your post, I was off for a bit…
My guess is that we need to improve the Wiki’s RC configuration page to make it more clear what the user needs to see to ensure the inputs are working correctly.
Perhaps we should also try and consolidate the output setup wiki pages (here and here)
Hi, i have the same situation. Can you elaborate what you mean by “if the rest isn’t just as documented” which rest is that? i have the feeling i have the same things missing. Or @rmackay9 can you explain.
I’ve had a quick look at the logs and a couple of things I’ve noticed:
ARMING_CHECKS = 0. i know checks can be annoying but they often display the cause of problems so I recommend returning this parameter back to 1.
RC input calibration hasn’t been done it seems
PILOT_STEER_TYPE = 1 which means two paddle input (wiki). Very few users want this type of input. Most want separate steering and throttle input which is “0”.
It’s best to include a dataflash log file so we can see what’s happening, without that it’s all guesswork. For example it’s possible that the vehicle has no safety switch in which case BRD_SAFETYENABLE should be set to zero. it’s possible the vehicle is disarmed so it’s not sending output. It’s all guesses without a log.
Arming Checks were disabled for the testing. Since all the motors are disconnected, nothing should go wrong. i now activated parts of it. For the final version I will all of them.
RC was calibrated, apparently i extracted the params beforehand.
I changed the steer to 0
You are right there is no savety switch, so i set BRD_SAVETYENABLE = 0
According to Mission planner, the rover was armed and in the raw sensor output I could also see the RC input, but the output is constant. I checked the code in git and all I could see was the check if the throttle_left and throttle_right is set in the Servo function in order to trigger the skid output. So I did not understand the other checks that @direktor was mentioning.
Please see attached also a short log where I arm and move the rc sticks around. I am not very familiar with analyzing logs, so I am very thankfull for your help.
It looks like you are using channel 3 and 4 on your transmitter but RCMAP is set for 1 and 3. It appears you are not arming and stuck in hold. I’ll recheck the parameter list but I think I also saw a RC TRIM problem on RC3. You may want to revisit the Wiki and take this step by step again.
I get your point, and it works now
Regarding the RCMAP, I put Roll to CH1 and Throttle to CH2.
I also redid the RC Calibration, just to be sure that the trim can not be an issue
Then I armed in HOLD, switched to manual and voila it worked.
