Home › Forums › Everything about everything else › Triflight 0.6 RC1
Tagged: cleanflight 2.0.0, drift, TriFlight, yaw
- This topic has 101 replies, 18 voices, and was last updated 7 years ago by Siclair.
-
AuthorPosts
-
5 November, 2016 at 17:23 #34586laukaParticipant
First release candidate for Triflight 0.6 has been released: https://github.com/lkaino/Triflight/releases/tag/0.6-rc1
Main driver was rebasing to Cleanflight v1.14.
I quickly LOS tested on my Baby, flew nice with the gyro notch filter enabled.
Please help testing so we can release the official version.
5 November, 2016 at 18:52 #34589neoxioParticipantHi lauka,
Thank you for new release 🙂 I will make some test tomorrow.
5 November, 2016 at 21:24 #34591swissfreekParticipantOh the timing. Rebuilding my Baby to switch back from a dRonin board to the F3FC tonight. Will test-fly tomorrow.
6 November, 2016 at 15:07 #34613laukaParticipantThanks! It’s getting too cold for my fingers and my batteries (-8C) to do any longer testing.
@swissfreek: how was the experience with dronin?
I’m thinking that I should base the Triflight stuff on betaflight now. If I just had the time.. Betaflight is stikk missing the parameter groups. Maybe after that is merged.
8 November, 2016 at 12:04 #34670TerjeModerator@lauka there is some really exciting stuff coming up in betaflight 3.1.0 shortly. Dshot seems to be the next big thing, supporting the 2048 step capabilities of the Aikon ESCs.
Finding weekends with nice weather for flying around Denmark is pretty hard at the moment so I won’t be able to test anything for a while. Well, hopefully Dshot and more exciting stuff will be matured when spring arrives 🙂
8 November, 2016 at 23:11 #34685swissfreekParticipantDRonin experience was alright. There were some weird behaviors that we couldn’t nail down with the baby (namely the servo seemed to wiggle with varying intensity based on throttle. Right around when I got everything going, macOS 12 came out, and something changed in the way dRonin was coded, which suddenly made me unable to build from source (im not exactly a wizard with that stuff). Then I had some inflight resets, and a Betaflight target was released for the board I was using (BrainFPV RE1), and I decided I’d rather have that in a quad where I can really play with it and leave the baby to a proven solution, hence back to the F3FC.
Now, as for 0.6RC1: it feels as though the tail wag issues I had with the 0.5RCs last summer is back. There is a strong very yaw kick on throttle punches, and the tail servo seems to oscillate. The oscillation seems to be directly proportional to throttle position (i.e., more throttle, more oscillation). I was concerned that maybe it was my tail servo acting up since I was having similar issues with dRonin, but that would be weird since I also rebuilt the servo with new gears while I rebuilt the baby with a new board. Reinstalling 0.5 made everything buttery smooth again, so it was definitely something new in 0.6. I imagine I could probably play with the yaw PIDs and see if that resolves some of the issues with the tail oscillation.
Weather here in NC USA is gorgeous at the moment (summers are the worst, 40°C easily, but fall and spring are nice 20°C days with plenty of sun and low wind). I don’t typically have time to fly during the week since I get home at sunset or after dark at this time of year, but we have a holiday weekend coming up so I should be able to do some testing between Friday and Monday. Let me know if there’s anything you’d like me to try. If this matters, my setup is a stock baby Tricopter with an HS1177 and a minimOSD (the Furious Piggy, really cool item on a small build!). Radio is a Taranis.
I would love to see Triflight based on Betaflight as that seems to be the future but what we have now is already pretty good!
9 November, 2016 at 00:49 #34686swissfreekParticipantI’ll test it on the V4 as well. Mini is currently out of commission since I broke the top plate/F3FC at a race and haven’t had a chance to rebuild it yet (plus the baby just flies so well!).
9 November, 2016 at 14:57 #34700LitterBugParticipantI Have 0.6 installed on my F3FC Baby Tri. Waiting on good weather and daylight to give it a fly. I’m going to start from scratch on the tuning since I hadn’t gotten too far anyhow.
Cheers!
LitterBug9 November, 2016 at 21:47 #34714billydParticipantI noticed some of the gyro commands in the CLI for .6 are slightly different. The stock defaults in .6 rc1 seem different than .5 with respect to these commands (wrt low pass filter). Is that correct or should they be changed to be more similar to those in .5? (for the baby)
10 November, 2016 at 01:37 #34717aoymParticipantI have a question about the new update. I haven’t tried it so I apologize. Anyway I noticed when going through cleanflight that I cannot access some portions of cleanflight in triflight. Will this be addressed in the update?
For example I want to play with voltage scaling because my voltage scaling is off. Here is what I get when I go to the power and battery tab http://imgur.com/a/O8FR7
10 November, 2016 at 06:20 #3472212 November, 2016 at 23:41 #34771LitterBugParticipantI tried running a tailtune on my baby tri with stock .6 PIDs and it about fried my motors in less than 30 seconds. Don’t currently have BB logging, but I’d bet there was some serious oscillations causing the heat. Going back to .5 to see if it flies like it used to. Also going to add a serial logger.
Cheers!
LitterBugEDIT: running motors manually from Motors tab does not result in hot motors.
13 November, 2016 at 01:25 #34776LitterBugParticipantFlashed back to 0.5, restored settings, and motors are running cool and smooth again. I had noticed the motors sounded really gritty when on 0.6.
Cheers!
LitterBugEDIT: I attached the Triflight settings I used for 0.5 and 0.6 for comparison. Basically just defaults plus rates and RX config are what I used in both cases.
Attachments:
13 November, 2016 at 06:11 #34788fomoco_44ParticipantHas anyone encountered an issue with the Blackbox logging in 0.6-rc1 on the F3FC mini tri board???
I had an openlog BB logger attached to UART2 working just fine in 0.5, but as soon as I flashed to 0.6-rc1 the logger stopped recording. It would power up fine and create a log but there was no data in it and the LED’s wouldn’t flash as per usual. I have tried it on both UART2 and UART3(2500000 baud) but no love.
If I plug the openlog into another flight controller it works just fine, so I’m pretty confident that the actually openlog hardware is fine. Could it be a system load issue??
I have only had a chance to fly it briefly since the upgrade, flies nice and feels smooth, plus tailtune completed quickly without issue.Also, when upgrading to 0.6-rc1 there is a F3FC and RCEXPLORERF3 option, which one should be used for the integrated F3FC board?
13 November, 2016 at 10:12 #34792laukaParticipantThank you all for testing! First of all, the F3FC hex file shouldn’t be in the release. RCEXPLORERF3 target should be used instead. Sorry about that. I guess quite many of you tested with the F3FC hex?
@swissfreek: did you rerun the bench tail tune?
@billyd: can you list which parameters are different for you?
@LitterBug: the default pid looptime is faster than on 0.5. That might affect it. Might also be that the filters ate nit working properly. I need to do some testing with BB.
There is a CPU load issue in 1.14 which I reported here: https://github.com/cleanflight/cleanflight/issues/2479
-
AuthorPosts
- The forum ‘Everything about everything else’ is closed to new topics and replies.