Tagged: iterm yaw pitch roll log
25 March, 2017 at 13:15 #36990
Okay back from the field, flew 7 batteries. I’m happy with the overall flight performance. In racing use it’s perfect.
However I found out a very interesting thing when doing freestyle flight with only yaw inputs at different throttle levels. I have noticed this in earlier Triflight releases, but haven’t really given any thought for it. When I want to do a full yaw spin without any roll input, the roll starts fighting the yaw so much that the copter refuses to yaw completely. It even starts yawing to the opposite direction. I got a log file where this happens multiple times, I uploaded it here. From the log we can see that there’s constant error in roll axis when the yaw stick goes to full deflection. The roll I term winds up and the front motors start forcing the copter to yaw in opposite direction, like in a quad. I haven’t completely wrapped my head around why this happens. The back propeller+motor is a lever that does apply force on roll axis, but I would have imagined that it would be in opposite direction. Need to study the log more. The problem is the roll I term in this case. There should be quite easy fix for this, in fact it’s the first ever feature I implemented to Triflight (check the first posts in the debugging thread) :). This feature didn’t make it to official Triflight, I didn’t see it necessary. Back then I suspended the roll I term based on yaw stick deflection and I think it should work pretty well here also.
I could release the 0.6 RC2 without this feature, but I kind of want to fix it before it. This doesn’t happen very often, only in freestyle flight. Usually we perform coordinated turns where we mix roll in with yaw. But you never know, it might help with the coordinated turns as well.25 March, 2017 at 16:06 #36992krkljanParticipant
Lauka, this is a very good news. In my opinion make release when you think Triflight is ready. Of course if there are testers willing to try then you release it but with strong warning notice. I hardly wait to get new triflight on my big tri. Nice work!!!29 March, 2017 at 13:14 #37034
Is there a support forum thread for triflight? I’d like to ask a few questions regarding setup…29 March, 2017 at 14:34 #37035swissfreekParticipant
There are many, but you could always start your own if you don’t see your question answered.29 March, 2017 at 16:22 #37036
I would like to ask if the Betaflight F3 FC will be supported any time soon? It has onboard current sensor, built in PDB and an OSD that id like to use!30 March, 2017 at 00:25 #37044swissfreekParticipant
I would speculate that that’s highly likely. @lkaino is currently in the process of adding the triflight features to BetaFlight. Since the BF3 is currently one of the more popular FCs for BetaFlight, I suspect it won’t be long after that that a target is built for it. Whether that’s anytime soon is a matter of your definition of “soon”. @lkaino is actively working on BetaFlight improvement and it sounds like he’s pretty close to having something ready for public consumption, but he’s a busy man, and very few people if any have tested these features yet, so it might be a little while until something stable is out, depending on your comfort level with beta firmwares.3 April, 2017 at 11:05 #37075
Sweeeeet ! – The sooner the Betaflight F3 board is supported the better as i’d like to use the OSD and the TriFlight tail tune.
I’ll send him a donation, see if that will speed it up a bit 😉6 April, 2017 at 09:50 #37105TerjeModerator
@lauka any hope of a TriFlight 0.6 Beta 2 based on Betaflight any time soon? I would really like to use the serial passthrough feature to setup my MinimOSD Micro 🙂
In short the passthrough feature lets you use the FC as a FTDI adapter, and using MWOSD Configurator you can easily setup and change your OSD layout without having to disconnect your MinimOSD from the FC.6 April, 2017 at 14:28 #37108LitterBugParticipant
@terje in theory, can’t we do that now in a round about way? Back up settings, Flash Betaflight on the board, do the passthrough, Flash back to TriFlight and restore settings.
LitterBug6 April, 2017 at 20:34 #37115
I have been super busy for the last two weeks. I did implement the last change and hopefully will have time to go test it on the weekend. If all goes well I will try to squeeze the release next week. Still have to figure out what to drop away from targets that don’t anymore fit the Triflight stuff.
Bare with me :).7 April, 2017 at 12:06 #37120TerjeModerator
Looking very much forward to the release, Lauka! 🙂8 April, 2017 at 18:11 #37138
I flew couple of batteries today, the change I made didn’t help much. I will leave it to the drawing biard for now.
Otherwise it flies well so I’ll try to make the release soon.8 April, 2017 at 23:24 #37139LarryTruParticipant
Excellent. There will always be stuff on the drawing board. I’m looking forward to the release too! 🙂
Larry24 April, 2017 at 10:45 #37386bhuismParticipant
Lets see if I can follow your thread of thought here. When the tail motor spins up, creates force, and it’s at 45 degrees for instance, it will create a yaw, pith and roll force. Wich must be compensated? or not? Am I correct in assuming that the PID controllers for pitch en roll are not changed at all? and because there is no stick input for these pitch en roll controllers, they will compensate automatically? or do we need some logic?
I will do some yaw spins today with a blackbox log with a betaflightf3.
ps: why not put the allready made fix for this in, if it works? to kill the roll I term when yaw input is given I mean.25 April, 2017 at 17:12 #37394
@bhuism: what you said is correct, except that I don’t think the angle of the servo affects the created roll force. Only way the tail motor affects roll axis is by the torque force, and the amount is relative to the speed of the motor.
The fix didn’t fix the problem, maybe helped a bit. I think changing the rotation of the front motors would help more, I will have to test it next time I’m going to fly.
- The forum ‘Everything about everything else’ is closed to new topics and replies.