RCMAP changes in for copter becareful with master

[Randy Mackay]

 

     Rob’s RCMAP fixes for copter have gone into master.  It’s a significant change affecting many parts of the code.  Please be extremely careful with master… and if you’re an RCMAP user, please give it a try (with props off at first!) and any feedback would be greatly appreciated.

 

    One thing is you must reboot the pixhawk after modifying the RCMAP parameters.

 

     By the way, two MP issues were found during testing:

                   https://github.com/diydrones/MissionPlanner/issues/883

                   https://github.com/diydrones/MissionPlanner/issues/884

 

-Randy

[Svein Alexander Frotjold]

Added reboot required to description in pr: https://github.com/diydrones/ardupilot/pull/2341

Is there any plans to expand this to ch5 and ch6 also?

as far as i can see this mostly resides in:

static void tuning() in tuning.pde

read_control_switch() in Switches.pde

but a structure change might be needed to make channel_flightmode and channel_tune separate from the aux switches.

for now i have flipped ch5 and ch6 in my branch to be able to have 3pos(3+2) switch on flightmode

This seems to work fine.

Also. i see the switch options for ch6 has been removed and this is now solely a tune channel.

Since i'm one of the few left with a DX7 this leaves me with only 1 aux channel for other stuff.

but it seems that i can use ch 6 for gimbal tilt so i can get by with that.

so the gimbal input channels should be left "un rcmapped" since it is selectable on the gimbal setup page.

[Robert Lefebvre]

Hi Svein, I'd actually like to make Ch5 and 6 more like Ch 8-12.  So that Tuning and Mode would be more like Aux Functions.  This would be easier and I think more logical than making Mode changing more like a flight control.  I'd like to keep the flight controls separate from the other usages.

[Svein Alexander Frotjold]

Sounds even better :)

Full flexibility.

[David Price]

Very cool :)

Previously I was compiling the code myself to change the channel order for the '3 position switch' function (I couldn't map them differently on the transmitter)

Thanks!

[Andy Piper]

I did a build of 3.3rc5 with this fix and then a fairly comprehensive test. Pretty much everything checks out and Michael O has fixed the MP issues so calibration was good (as an aside the fixes he has made to MP also means that if you use the old code you cannot calibrate the RC with the new version of MP when using mapped channels - MP keeps saying RC not calibrated. I believe this to be a good thing!). I ran in AltHold, Loiter, PosHold and did a couple of simple auto missions - everything was good. Control appeared identical to what I was experiencing using the spektrum port on the pixhawk.

Good job Rob!

There was one small problem - and I experienced this also with my version of the fix as well and could not figure out what the issue was, so it may be subtle - MOT_SPIN_ARMED does not seem to work (or at least gives different behaviour to before). I have it set to 90 and the props never spin on arming - tried with different batteries, multiple arms - never spun. On top of this when I raise the throttle, the bite is fairly aggressive - I go from 0 to quite fast, there is no gradual spin up. A small thing, but it would be nice if it worked. It works without mapping and using the spektrum port.

[Andy Piper]

So I've been running with this for a while and I have one weird problem. Flying on a second battery I get a radio failsafe triggered. I've had this happen twice to me now. But the failsafe isn't the usual loss of signal - it's code "2" which appears to be because a value couldn't be read for 500ms. The outage lasts about a second and clears up immediately after the failsafe has triggered. The copter wasn't that far away from me (50m maybe?) and I have flown out to 150m with no issue, so I don't think this is a radio interference issue. Now granted this is a lemonrx and I guess the most likely explanation is some issue with the RX, but I raise it here in case it rings any bells.

andy