Testing – Triflight rebased on 3.2.2 Betaflight

[Terje]

@lauka would it be possible for you to compile targets for the Holybro Kakute AIO v2, Omnibus F4SD and Omnibus F7V2?

[Nathan116]

Being that this has been out for like 6 months, is this BF 3.2.2 considered stable or no?

[Kevin_Erik]

Yes but Odds are we may be due to an update considering the performance enhancements that are now in 3.4 Betaflight.

[FPVienna]

I just decided to try this firmware (triflight_bf_3.2.2_v2_NAZE) and noticed a tiny difference setup-wise between this firmware, and triflight 0.7 Beta3.

So, just a tiny warning upgrading their NAZE to this release – be sure your motors are on ports 3 to 5 (using resources in B0*) and your servo is on port 1 (using resource A08). If set up otherwise, your motors might stop working correctly when upgrading.

Have fun! 🙂

[Joseph]

Hello Triflight users,

I have just finished a build of my new design with a Airbot Omnibus F4 – V5 FC, Airbot Wraith 32bit DShot1200 30Amp MiniV2 ESC’s, RC Explorer PDB, Digital Servo HK 210MG, Cobra 2206 – 2100Kv motors with 6 x 45 Tri-blade HQ Props and 3S – 1400mAh LiPo.

Have managed to get the setup flying with Betaflight 3.5 and using stock PID’s.
Let me say, flight characteristics on first maiden, wow, crisp, sharp and the only issue I have experienced is a slight left to right tail wag in a hover which I hope to correct with a bit of tuning on the Yaw, maybe lower the I gain a little. I have no feedback.

Other than that, it is sweet. So far flying this setup the other day in 30Kph gusty wind the batteries had 40% left in them after 5 minutes of flying, ESC’s cool, Motors cool and Battery cool.

Base frame is 4mm carbon, side frames 2mm carbon. Tail Mech, my own design and make, combination of Nylon and Aluminium.

I have attached a couple of pics so you can see. I can recommend giving Betaflight 3.5 a try as it seems to be doing some magical things. Regard’s, JosephR.

Attachments:

1. 

   IMG_3584_web.jpg
   

2. 

   IMG_3586_web.jpg
   

[Kevin_Erik]

@Joseph

Would you be so kind as to post your Diff and maybe a few screenshots specific to your Servo setup?

[fotoamg]

hi! I have now the local build environment setup (build modified inav and beta easily).
Can someone give me a link to this “Triflight rebased on 3.2.2 Betaflight” or rebased to some newer Beta repo/fork in github to clone?
Thanks!

[Kevin_Erik]

Do a google search for Triflight and Github. Should pop right up.

[Joseph]

Some more information about my build.

Dry weight = 433 grams
Flying with 3S 1400mAh battery = 577 grams.

[Joseph]

Here is a Dump of my Betaflight 3.5 settings, see attached Text File.

Attachments:

1. TRI230-Dump-BF3_5.txt

[Joseph]

Kevin,

Attached are a few images of my Tail Mechanism, the Cap screw is 4mm diameter and has a long shaft that is not threaded. I have also taped a 4mm thread into the servo output shaft. Material is Nylon for it’s toughness.

If you have any comments about Tail Wag and what causes it and the best way to get rid of it, please let me know. Thank you.

Hope these help. Regards, JosephR.

Attachments:

1. 

   IMG_3589_web.jpg
   

2. 

   IMG_3590_web.jpg
   

[Kevin_Erik]

Right off, i saw that your using a tilt mech thats similar to David’s. This method, while it does work, it places the motors mass well above the Servos axis of rotation. This means that the Servo now has less torque and a slower response. Furthermore, the tail motors spinning mass Yaw’ing to one side or the other can cause the entire copter to Roll in the other direction. It being mounted higher than the Servo amplifies this effect.

Please keep in mind that this single motor controls both Yaw and Pitch for the entire copter. Thus a slower Tail tilt mech responds poorly and can induce oscilations on both Axis.

Suggestions

1) Try the Impossible Tilt Mech from Thingiverse.

2) Use a strong screw for your Tail Mech. Aluminum bends easily so i am swapping all of my Tail Mech screws to Titanium.

3) Wrap the screw with a thin layer of Teflon tape. The tape has an extreamly low resistance coefficient and should help prevent binding. Better yet, dust and dirt don’t stick to it.

4) Don’t use any form of soft mounting for the tail motor as this will allow the motor to flex. This could result in a dampened response which may not be desirable.

[fotoamg]

imho tail wag is for me in the ground before I take off (that wag is not in cleanflight 0.9 so I am using that for now) and could be removed somehow to not compensate yaw as it cannot do it in the ground so when there is very low throttle… maybe with some PWM value limit and threshold from some setting or calculated from hover throttle but the yaw acc parameter makes the opposite increases yaw strength at low throttle

[fotoamg]

“Do a google search for Triflight and Github. Should pop right up.”
I searched but only the Triflight repo is what I found which is not the repo where Lauka rebased it to 3.2.2.
I only found the official Triflight repo:
https://github.com/lkaino/Triflight/releases

What to search on than?

[Joseph]

Hello Kevin,

Thank you for your comments and suggestions.

Regarding my tail Mech design, just a few comments. Yes technically you are correct in that my design is, you may say top heavy, however having used this type of design now for 5 years on a number of Tricopters, I have never had a servo fail and not had a circumstance where the servo could not keep up with the input I have required when flying around. With this build, I have lowered the motor mount as much as possible so to minimise the offset. The gimble type design of the Impossible Tilt Mech whilst is low and yes more balanced is too low and for my frame design does not provide enough prop clearance over the frame, especially when the tail mech tilts.

The pivot bolt I use is a 4mm steel socket head cap screw and it runs through nylon which provides good support and to some degree is self lubricating, although a squirt of silicon lubricant does help. The aluminium plate that runs under the motor provides a place where the servo horn can be attached to transmit the servo movement to the motor mount. The pivot bolt goes all the way through this and screws into the threaded hole in the output shaft of the servo.

Front motors are soft mount and agree, don’t soft mount tail motor.

Did some tuning of the Yaw last night, lowered the I gain a bit and tail left to right wag movement is less and only slight now in hover so no real need to do much more as it really flies great.

I hope this conversation is a benefit to everyone, it will be interesting to see if we can see a version of Triflight based on Betaflight 3.5 as it certainly has some great features and performs very well. Regard’s JosephR.

[Author]

Posts