Home › Forums › Everything about the Mini Tricopter › Debugging the tricopter mini racer
Tagged: F4 flight controller
- This topic has 939 replies, 82 voices, and was last updated 7 years, 2 months ago by lauka.
-
AuthorPosts
-
31 October, 2015 at 19:57 #22008Brighton till I flyParticipant
I didn’t do the best tune on mine triflight 3 feels better and less wiggly than 2 but in forward flight My mini yaws to the right badly un-flyable in a race :C
2 November, 2015 at 02:49 #22046LeoTheHumanParticipantUnfortunately, there was a very heavy fog in London this weekend, so I was unable to fly. Yaw issue is most likely caused by yaw TPA, which is something lauka plans to remove for the next version, and less than optimal tail settings, which he was tuning on Saturday morning. But he’s very busy this weekend, as far as I know, so I would probably expect the triflight release closer to the next weekend. Weather around here is expected to get better by Thursday, I will do my best to get the next version by that time, even if I have to code it myself. (:
2 November, 2015 at 04:51 #22050storminParticipant@lauka and @LeoTheHuman can you share the latest release or the optimizations via CLI? Looks like I may have decent weather for my daze off this coming Wednesday thru Saturday. All my minis have some form of alpha/beta/rc FW installed – I enjoy trying out bleeding edge goodness!
2 November, 2015 at 13:02 #22056laukaParticipantNew version will be released this week. I was very happy how it flew this weekend, had time to test and tune for 5 batteries. It was very windy which might also be a good thing for the tuning :).
It will contain:
– remove the TPA affect on I term
– new defaultsIf there was time to test I would also merge the latest Betaflight, but next testing time is on weekend. It’s already quite dark outside after I leave work at 4 PM :).
3 November, 2015 at 18:46 #22092laukaParticipantTriflight 0.3 is available here: https://github.com/lkaino/Triflight/releases/tag/0.3.
Changes compared to 0.2:
– Major bug fixes to yaw axis motor control
– Float operations converted to integer (faster cycle from gyro input to motor output)
– New default tune for RCExplorer mini tricopter
– TPA no longer affects I termAs before, any feedback is welcome. It would be helpful if someone with a blackbox logger could record the servo movement (camera facing backwards) while doing hard yaw maneuvers. The servo speed is now set to 300 by default, I don’t know if it’s too little or too much for a 5V Bluebird servo.
3 November, 2015 at 18:49 #22093laukaParticipantWhat’s in the roadmap for future:
– Merge to latest Betaflight
– Implement YAW TPA for PID controller 1 and default tune (Multiwii rewrite, the integer controller)
– Investigate occasional bounce backs on hard yaw maneuvers
– Changes based on feedback
– Code cleanups4 November, 2015 at 03:32 #22104Brian BlackParticipant@Norman, I have been having a really hard time lately and haven’t flown in so long I can’t remember…
Daze means not clear-headed, (maybe from a blow to the head) or a state of bemusement (laughter). Days is more than one day. I like your pun either way, intended or not.
🙂
4 November, 2015 at 04:31 #22106storminParticipant@Brian Black Ha – yeah I was intentionally trying to be punny. @lauka very cool to see Triflight 0.3 available I’ll install it in the morning and give her a test fly.
4 November, 2015 at 12:11 #22114biggestRCEfanParticipantHad trouble trimming my tri. It was drifting backwards on me and no amount of acc_pitch_trim would correct the problem. Then I discovered that the accel calibration was the issue. I had the tri sitting on a level surface, but due to the vertical panels of te canopy extending below the bottom plate at the nose, it was actually sitting nose-up by 1.6deg! This was sufficient to cause my trim issue.
Lesson: Make sure the Naze frame is level when calibrating!
4 November, 2015 at 14:24 #22115LeoTheHumanParticipantWell, that is quite predictable. What I usually do when I start on a new frame is I calibrate the accelerometer, take off, look where it drifts, land and use stick commands to trim at least 1 degree opposite of the drift on each axis. Repeat as needed. As soon as I go over and it starts drifting the other way I divide the trim by 2 and repeat that procedure till my trim gets down to 1 decidegree. That’s the best you can do and it should be good enough. It doesn’t take a lot of time (like, half a minute to a minute) and only needs to be done once. Make sure to mount the FC firmly before that, as it’s very sensitive to FC position and if it’s mounted with a zip-ties or just hanging around on the bolts without being strapped by nuts on top, it tends to move enough to make the effort worthless.
It’s unrealistic to expect perfect leveling of the board, so trimming the accelerometer is a procedure you have to learn and it’s very simple to do. I even started to enjoy it after a few tough builds! (:4 November, 2015 at 18:18 #22122jonniParticipantHas someone tried the Triflight 0.3 (or an earlier version) on the full size tricopter? I experience some yaw drifting, especially when flying backwards with my full size tricopter and would like to see if changing to Triflight makes any difference.
Which settings would you recommend when running the Triflight firmware on the full size tricopter?
4 November, 2015 at 21:30 #22143hrrrParticipantis the servo lowpass filter recommended on triflight 0.3 for the mini?
I saw it is disabled per default in triflight, but enabled to 120 in David’s settings…Thanks!
5 November, 2015 at 08:33 #22148laukaParticipant@Jonni:
I haven’t yet tried Triflight on V3. It’s in the plans. Biggest concern is the Betaflight 1000 us terget looptime. Many people have bells and whistles attached to V3 like GPS, LEDs, etc that take execution time. If you do test it on V3, check the looptime value on configurator. Otherwise it should work fine after some PID tuning.@hrrr:
I don’t see much benefit from enabling the lowpass filter for the tail servo output. Would be nice to hear from @David the reason for the filtering. Looking at the blackbox logs I haven’t seen much noise in the servo output after PID controller and mixer.5 November, 2015 at 09:51 #22150RCExplorer – DavidKeymasterI threw the filter on there during early testing and it helped with the “feel” back then. But I’m sure it’s not necessary if Lauka disabled it 🙂
5 November, 2015 at 11:20 #22156GAntonjoModerator@lkaino: I know you are busy with more than triflight, but have you seen this issue from cleanflight? From what I understand you should pull in this solution in order to make sure (Windows) users don’t have to disconnect and connect again manually after each save&reboot. That is, it is not an error in CleanFlight configurator or Chrome, even if the latest Chrome update caused this error to appear.
-
AuthorPosts
- The forum ‘Everything about the Mini Tricopter’ is closed to new topics and replies.