RC: improve detection of weak RC protocols

[tridge]

This is to fix an issue with detection of DSM on a vehicle that actually has SBUS. It just increases the number of frames required for DSM detection to 10

@MichelleRos this PR is because of the issue on your quad yesterday. It seems that a soft reboot when you are on batter power and connected via SiK radio sometimes detection DSM instead of SBUS. Can you please first reproduce the issue with current master (I think it happened maybe 1 time in 3 on soft reboot?) then see if this PR fixes it.
Thanks!

[MichelleRos]

I tried reproducing the detection error and it detected SBUS every time... No idea why it detected wrong yesterday.
(I rebooted it about 15 times and couldn't get a wrong detection.)

[tridge]

I tried reproducing the detection error and it detected SBUS every time

were you on battery power, with SiK radio telem and no USB connection?

[MichelleRos]

were you on battery power, with SiK radio telem and no USB connection?

Yes. Even connected via MAVProxy not Mission Planner in case there's any difference there. I also tried moving the quad further away (to another room) in case extra noise from the distance or something makes that difference.

Just realised that I didn't test whether switch positions could affect it - like perhaps it's if the channels are at certain values that it makes the SBUS look more like DSM?

[tridge]

Just realised that I didn't test whether switch positions could affect it - like perhaps it's if the channels are at certain values that it makes the SBUS look more like DSM?

good thought, that could well be it!

[MichelleRos]

Tried again with different RC inputs (of the values that it could've been that day), but still no mis-detection.
Unfortunately there's two channels that could've been various values (throttle may not have been quite at the bottom and the one slider that I just realised outputs to RC that I never use 'cause it is to easily to accidentally knock), so it's a bit hard as well to test all possibilities.

[tridge]

@MichelleRos I was connected at the time I requested the reboot, and recorded this on SBUS just before the reboot:

2023-02-12 13:34:04.851 RC_CHANNELS {time_boot_ms : 400211, chancount : 16, chan1_raw : 1493, chan2_raw : 1495, chan3_raw : 982, chan4_raw : 1496, chan5_raw : 1408, chan6_raw : 982, chan7_raw : 982, chan8_raw : 982, chan9_raw : 2006, chan10_raw : 1495, chan11_raw : 1495, chan12_raw : 1495, chan13_raw : 1495, chan14_raw : 1495, chan15_raw : 1495, chan16_raw : 1495, chan17_raw : 0, chan18_raw : 0, rssi : 0}

[tridge]

@andyp1per can you test with DSM ?

[tridge]

@MichelleRos has shown this does NOT fix the issue, new version of the PR soon

[peterbarker]

@MichelleRos has shown this does NOT fix the issue, new version of the PR soon

Real Soon Now?

@tridge can you recall if you actually had a replacement branch for these patches?

[MichelleRos]

Real Soon Now?

I thought we'd mostly decided that since this isn't a flight risk since we no longer redetect later, we were just going with the plan that if this happens we tell people to set the RC_PROTOCOLS parameter?