Ok - now I'm confused...

[CarolineTyler]

Decided to change the flight computer from CC3D to a NAZE32 so I can run cleanflight and use triflight software from RCExplorer. My tricopter is very closely based on his super design.
All looked great until it came to tuning the rear servo......1st it won't move unless armed, so disconnefted the motors for safety and now when it's in the centre, I get massive jitters, move it left or right and its super smooth! What am I missing here? The rear servo is the same as RExplorer recommends only without the extra feedback wire.

 

[PsyBorg]

I am not the tri expert by any means and I am sure Litterbug will chime in when he sees this as he is the resident crazy person with these. I have tried to keep up with what they do with tris best I can so I will try and give you some ideas to think about.

When working on the bench for set up I would start by disabling the gyros (any auto level function) as we can never get things perfectly level on the bench and they can inject more issues trying to find a happy spot it thinks is level. What I think is going on or at least part of that is when you go to center stick on rudder the FC tris to make a correction to keep it level. when it gets no change it tries again to correct. With the brutish movement it makes with that tail servo it really shakes the frame and FC making it adjust harder and faster to oppose that movement. Thus it creates a fast loop right at center stick. Once you go away from center it no longer tries to seek level as it thinks it is turning from a manual input.

Something else to think of is what type of gimbals you have on your radio and what condition they are in. If you have a lot of drift from wear on analog gimbals potentiometers that can make it hunt as well because of electrical slop. If you look at your radio where you can adjust center and end points and watch the receiver tab you will see the numbers at center will drift up and down a wee bit. If that drift is more then 2 points in either direction (1498 -1502) you may have to add some dead band so the FC ignores that drift for that axis.

Those things are probably not the cure for what is going on but may help to track down what is going on by eliminating outside factors we can control during the troubleshooting process.

 

[CarolineTyler]

The radio is new, centre point set at 1500, low 1000, high 2000 and a jitter of no more than one. So it's not the radio, I will do some more googling on this, as I've already found some stuff on resource mapping needing to be updated via the CLI. Not going to let this thing beat me!!!
Thanks for your suggestions

 

[PsyBorg]

Thats the key kiddo. Never give up. The solution is out there some where. Like I said there are pros here with that whole tri thing. I just try and keep help requests in the loop so people who can help see them if I am unable to resolve things.

 

[LitterBug]

A: I'm going to start off with my "opinion"... If you are using the latest TriFlight 0.7, Roll back to the last released 0.5 version. The 0.7 version has some really bad yaw waggles that seem to plague it. One of the dRonin developers that I have been working with on triflight integration has plotted out some simulations and found that the yaw gain at low throttle is to blame.

B: Add tailtune flight mode so you can easily set the servo end points. Enable tailtune mode when disarmed. Push up on the pitch stick. Use yaw left/right to set the center point of the servo so the prop is level. Push roll stick to the right and servo should go to the right endpoint. Use yaw left/right to set the servo so the prop is at 40 degrees. Push roll stick to the left and servo should go to the left endpoint. Rinse and repeat setting prop to 40 degrees. Push up on pitch and it should return back to center.

C: Props on and carefully try a test hover outside....

D: report back. :-D

E: if you are still having issues, add a code section with a "diff" from CLI.

Cheers!
LitterBug

PS: Oh yeah crap, 0.5 requires the old old old cleanflight configurator... (this is why I use dRonin LOL)

 

[CarolineTyler]

I went to download dRonin but the PC warned of malware in it so didn't install !

 

[PsyBorg]

I am pretty sure Dronin is safe as long as you got it direct from the devs. Litterbug has been beta testing for the developers for over a year now. Even though he is a madman for building and flying so many tris he is trustworthy and will not steer you wrong. More then likely its showing as questionable because of drivers. Hang out a bit and I am sure LB will be in to tell you what is going on.

 

[LitterBug]

I went to download dRonin but the PC warned of malware in it so didn't install !

That is a false positive. I'm running a non-stock branch that has @Lauka's Tri-Flight tail routines merged in. Support for FCs with F1 processors is limited as is the case with most firmwares these days.

You should still be able to do the "tailtune" part even with TriFlight 0.7. If that isn't working, then there is something not configured correctly. Can you provide a "diff" from cli?

Cheers!
LitterBug

EDIT: PS, I'm going to check with my dRonin dev buddy to see if his branch currently supports the Naze32. He has a naze32 Pro which is slightly different....
EDIT2: If you PM me with the OS and what anti-malware you are running that is giving the warning, I'll forward the info to the devs so they can add it to the list of vendors to be updated.

 

[CarolineTyler]

Installed dRonin Artific, and no more tail servo wagging!!!!!

 

[cranialrectosis]

Nice. I gotta try dRonin on a tri-copter.

Does the autotune feature work or is this a triflight variant with the tail tune function?

 

[CarolineTyler]

Nice. I gotta try dRonin on a tri-copter.

Does the autotune feature work or is this a triflight variant with the tail tune function?

Autotune in in it, I will be trying it tomorrow, weather permitting!!
This is a superseded version as my Naze32 is not supported in the latest release

Caroline

 

[PsyBorg]

Well done! Whats the chances of seeing video of the tail tuning process from a new pilots perspective about what you expect and what actually happens?

 

[jhitesma]

triflight hasn't made it into the official dRonin releases yet. the developer working on it releases his own builds with it - I think he sometimes posts them on David's forum at https://rcexplorer.se

Stock dRonin autotune does work on tricopters though - it just doesn't have the special tri-only improvements by default yet.

Artifice is the last release of dRonin that still runs on F1 boards like the naze, and autotune has been improved substantially since that release (due in large part to not being restricted by the limitations imposed by F1 boards) so you'd probably get even better results with a F3 or F4 based board.

 

[CarolineTyler]

I tried the autotune this morning. Think it was a little more windy than it should have been. Broke a prop and a prop support at the end, new one being printed as I write this

 

[LitterBug]

I've been playing with the Tricopter branch of dRonin for quite a while now. I pull the branch down from the dev's github and compile it myself. Tri's do fly well on the official dRonin releases, but there is a noticeable improvement in the yaw behavior with the TriFlight tail routines merged, especially with more spirited flying. The branch also adds dRonin support for the RCExplorer F3FC racing FC. The nice thing about that, is it makes it possible to compare the BetaFlight based TriFlight to dRonin based TriFlight on the exact same hardware.

Cheers!
LitterBug

 

[CarolineTyler]

Sadly, the auto failed.....it was still doing its wriggling around when I had to land due to dogs...

 

[cranialrectosis]

Tricopters may not be as tasty as a rubber ball, but they are crunchy and spark when you chew them!

 

[LitterBug]

I tried the autotune this morning. Think it was a little more windy than it should have been. Broke a prop and a prop support at the end, new one being printed as I write this

Make sure you calibrate the ESC endpoints. It looks like the tail motor was not responding the same as the other motors. Also, Click the box "motors spin at neutral...." on the output tab, and set the neutral value the same on all three motors to a value where the motors spin smoothly.

Others viewing this thread did not see our PM where we were discussing the Naze32. The standard naze32 has an F1 processor which went out of support on dRonin after the artifice release. There was a very limited run of naze32PRO boards that had F3 processors which would have been supported on newer dRonin releases and the Branch with the tri routines merged.

Cheers!
LitterBug

 

[PsyBorg]

Sadly, the auto failed.....it was still doing its wriggling around when I had to land due to dogs...
View attachment 106500

Don't you love it when people think their the pets they love are loved by everyone and let them invade others space uncontrolled. I got a friend here where I live whos Jack Russel I love but man not when I am trying to fly. I can not count how many times I have been under to goggles and he has walked up and let the dog jump on me or up on the table where I fly only to get all tangled in my gear or knock stuff off the table.

Nice thing is you can clear that tune attempt and start over from what I have learned reading LB's posts. Better luck on the next attempt.

 

[CarolineTyler]

All repaired!!!
Love the ability to be able to just print out a replacement part rather than waiting for it to be shipped all the way from China