Tricopter LR – Firmware (Triflight, iNav)?

[fgsfds1234]

this is actually what I came up with. in reading the page on the esc (rdq 32 bit esc) they mention using a certain version of betaflight or newer, so I just flashed betaflight and it worked just fine (on one of the four versions of blheli at least) there goes half a day of messing with drivers and pulling out hair. at least I took notes for servo setup. hiding behind a chair I hovered it for 3 seconds, so excited hopefully I can fly it tomorrow after I loc tite all the screws. it looks like there’s enough stuff on OSD for now, that’s a problem for another day.

[billyd]

OSD is easy it’s just trial and error moving stuff around adding or removing. Located in the system tab.

Unfortunately my osd died and I only got a couple of flights in. Now my tri is back in pieces as I wait for the new fc to arrive. I am contemplating contacting holybro and complaining. I did nothing to this board to cause it to fail.

[mrmattbill]

Anyone else have a LR running betaflight? I just got mine to actually fly today, Im running all Davids recommended hardware + Crossfire, GPS, and smart audio. Now I just need to do alot of work on the PID tuning. I spent probably 30 minutes today just messing with the yaw enough to get it to not wag back and forth, but it still flies pretty poorly.

Anyone have some known good PIDs I could start with?

[ineedtubesIT]

Anyone else have a LR running betaflight? I just got mine to actually fly today, Im running all Davids recommended hardware + Crossfire, GPS, and smart audio. Now I just need to do alot of work on the PID tuning. I spent probably 30 minutes today just messing with the yaw enough to get it to not wag back and forth, but it still flies pretty poorly.

Anyone have some known good PIDs I could start with?

I’m trying to use betaflight,so far it’s unusable. The tail starts oscillating more and more,I doubt it would stay airborne more than 5 seconds if I took my hand off of it.

Are you having the same problems?

Changing PIDs doesn’t seem to help.

[swissfreek]

I would think triflight would be more stable than that, surprising that it’s so rough out of the gate, but the version of Betaflight that Lauka used is so old at this point, who knows. I tried iNav for about an hour, messed with the PIDs for a while and never got it to fly very smooth or stable. I tried dRonin and had much better results so I’m sticking with that for now.

[ineedtubesIT]

Yeah I would use dronin or triflight but my fc (matek f722-se) is not supported.
I’ve been trying for a couple hours to fix this tail wobble but no luck so far..

The fact is that it’s so rough,it literally yaws and wobbles 20-30 degree on the horizontal axis.

[jihlein]

Not having used betaflight, I’ve no idea where the gains should be. You can try this method FWIW.

1)Zero the yaw I and D terms. Reduce the yaw P until it flies without oscillating. It will feel very loose and not hold a heading well at low P only gains, so be prepared. Now starting increasing the P gain until it starts to oscillate. Note the P value, and then set P to ~85% of that value.
2)Now start increasing I until it oscillates. Note the I value, and then set I to ~85% of that value. The tail should behave fairly well at this point.
3)I typically leave D at 0 for servo controlled axes, you’ll have to experiment with that if desired.

There’s a bunch of betaflight filtering options that could be getting into the mix here too, I’d try to make the initial setup as simple as possible.

[jihlein]

” I tried iNav for about an hour, messed with the PIDs for a while and never got it to fly very smooth or stable.”

Another note from the FWIW column……

I have a Fortis Airframes Titan tricopter, that I fly with an admittedly older version of iNav, but it flies absolutely perfect using the default tricopter gains. It flies so well I’ve never upgraded iNav to the newer version(s). This is also on a custom target I created for the AQ32 boards I have.

I wonder if something has changed along the way in the newer iNav builds that makes tuning a tricopter hard?

I’ve also looked at putting triflight into iNav, it’s quite an effort that I don’t have time for right now…..

[swissfreek]

It might not be hard, I’ll be the first to admit I suck at tuning. I was also still fighting the GPS/video gremlins *and* on a work trip so I didn’t have all my stuff and was in an unfamiliar location and didn’t want to crash, etc. etc.

Now that I’ve got it flying well with dRonin and know “what right looks like”, and have my video and GPS issues sorted, I may give it another go using your tuning suggestion above (what I call the “Stingersswarm method” from back in the day when tuning was mandatory even though I know he said he didn’t invent it). I use inav in all my flying wings so I’m quite comfortable with configuring it and using it, and it’s development is much more active than dRonin. I deliberately set my LR up so I can bounce back and forth without changing wiring so I could keep trying. If I can’t get it flying as nice as dRonin, though, forget it.

I’d actually kind of prefer if I could get it working without triflight, since like I said iNav development is pretty lively and adding meaningful new features regularly, and then you’d get stuck in the loop that lauka did with betaflight (I think), trying to keep up with new additions constantly.

[jihlein]

For the purpose that the triLR is intended, I’m not convinced triflight buys you much. I’ll eventually fly without it to see.

That said, I’ll eventually try iNav too. But I’m going to be on the road quite a bit for the next few months, so it probably won’t happen for a while. I still only have one flight on my LR…… 🙁

That reminds me, I need to submit an iNav PR for the Kakute tricopter configuration.

[ineedtubesIT]

Thanks for the little guide!

I’ll try to tune the tricopter again maybe tomorrow or some other week.

Hoping to report back with a positive result 😀

— Almost forgot,my last attempt at tuning was lowering the pid loop and gyro update frequency,and it seemed to behave better.
Is it possible that lowering it makes things better? Right now I’ve set 4Khz on gyro and 1Khz on pid loop.

[swissfreek]

@jihlein you and me both, man. I’ve got a total of two flights that I’d consider anything more than a test hover. I’m stuck here for the month of June, I’ll be home for a week and then I’m gone again until mid-August. I brought my LR with me but you’re probably well aware of the limitations I have in finding a place to fly (similar situation to your travels)… If I do get a chance to fly, which is unlikely, I for sure won’t be doing anything that could cost me my LR while on the road. So here I am on my phone reading about developments and chewing on it in my head and making plans, haha.

[mrmattbill]

I played with it for about an hour today trying to PID tune it. I did get to a place where its “flyable” but its still not anywhere close to stable. For Yaw right now I have p-80 i-5 d-20, I know that sounds weird, but its proving to be the best ive got it yet. Raising the I anymore at all seems to just increase the wobble. I had to introduce some D to get it to stop wagging though which is not typical for yaw I know.

Im open to any suggestions.

Also, I just realized today that I did use Davids feedback servo and wire it in, but I know he is running Dronin and not BF. Does anyone know if there is a way to utilize that with BF?

[swissfreek]

With BF yes. INav no. Look for “triflight” in these forums. But be warned, I believe it’s based on 3.1.7 or something like that from late 2017.

[mrmattbill]

Yeah, I’ve been down that road, but the Kakute f4v2 isn’t supported by triflight, most likely because as you point out it doesn’t seem to have been update in more than 2 years, which makes me nervous anyhow.

[Author]

Posts