Revs' Talon Tri (re)Build Thread

[Revs]

I';ve finally got a definitive reply from Adrian Round over at DIYdrones..

http://diydrones.com/forum/topics/where-has-the-mot-sv-speed-option-gone?xg_source=activity

Looks like the ch7&8 servo outputs default to 50Hz and aren';t adjustable without altering the ESC outputs.

Yesterday I fitted a Hitech HS-65MG servo to the tri which I picked up from my local shop as the TH servo hasn';t landed yet. Awesome little servo with great quality gearing and very little backlash. The yaw movement is way smoother now  ::)

The only problem now is that I think the copter is just too rigid. The fast/harsh vibration when yawing is all but gone, but the rear boom kinda oscillates slightly all the time like it';s bouncing on the slight flex in the arm. The servo isn';t moving so it';s not caused by that. It doesn';t really affect the flight but I';d like to to get rid if it if possible. I think some kinda of AV damping where the rear arm connects to the body is in order but I haven';t thought of a good way to do it yet. Maybe the arms being so long is now the issue..

[Revs]

I';ve tried to get a vid of what';s happening, but it';s very had to see it on the vid...

https://www.youtube.com/watch?v=On7ACdyPL0A

[Revs]

Wuhoo I think I';ve sorted it!  ::)

In my effort to make the tri as ';good'; as possible, I made sure I thread locked and tightened everything right up so nothing had any play. The arms were slightly stiff to move when folding and the rear boom solid. My thinking was that a solid craft made for a solid flight. Wrong.

The copter was so rigid it actually had a resonance frequency Once in the air, all the energy from the motors correcting and the tail servo stepping was bouncing around the frame. Don';t get me wrong, the yaw assembly has to be as solid as you can get it so the motor can only freely move in one axis. Any slop at all in the bearings and you get vibes as the motor flops around. But the arms need to be able to move a bit so the energy can dissipate.

All I did was loosen the pivot bolts on the front arms a quarter turn so the arms swung freely when being folded. Then loosened one of the bolts holding the rear boom to the body a tiny bit so there is the slightest bit of play. Boom! Totally smooth, vibe and oscillation free flight

The rear boom bolt took some moving due to the thread lock, so there';s no chance of it coming loose. Not ideal I know but without designing some sort of AV setup for it I can';t think of another way round. Plus those screws are the countersunk type, so they don';t go totally sloppy when a little loose. They allow for adjustment.

So, for the first time ever since first building the Talon tri, it';s flying as I want it and as it should. These batteries are great, too. 16min flight with plenty of stick input (loving PosHold mode  ::)).

[guest325]

Wuhoo I think I';ve sorted it!  ::)

The copter was so rigid it actually had a resonance frequency Once in the air, all the energy from the motors correcting and the tail servo stepping was bouncing around the frame. Don';t get me wrong, the yaw assembly has to be as solid as you can get it so the motor can only freely move in one axis. Any slop at all in the bearings and you get vibes as the motor flops around. But the arms need to be able to move a bit so the energy can dissipate.

Great to hear that you';ve got it sorted!
Yes making it too rigid can turn it into a tuning fork - my Conduit tricopter did that, and ultimately led to its demise 

[Hands0n]

Great stuff Revs - good to read you';ve got it sussed.    Look forward to the vids etc.   

Shame I missed that delicious GPS sidebar conversation, but I won';t re-open that now 

I';ve yet to lay hands on my Talon tricopter other than to fiddle with it on the bench a bit.  Not sure what FC to mount in it. I have a "designated" APM 2.6 for it - but also a KK2.1HC that is looking for something to play in.  Seriously thinking APM because if the Talon is right for me then as a folding multi then its going to be ideal for me when I travel on my holidays.  Maybe.

[Revs]

You should dig it out It really is a nice bit of kit. It just has the odd issue which, while being a bit of a PITA, shouldn';t be enough to put you off. FC wise, I';m not a fan of the KK unless it';s purely for acro use as the self-level is pretty poo IMO. I love the APM though, such a capable bit of kit, especially for the money, with massive support. Now improved with AC3.2!

Here are a few Talon Tri tips from what I';ve learned;
- Use semi-setting thread-lock everywhere. Delta D77 is perfect
- Shim the bolt in the yaw assy to eliminate play in the motor mount
- Don';t over-tighten the pivot bolts/nuts for the arms
- Use low-profile motors, preferably with CF T-mount props
- Use a good analogue servo for APM (Hitec HS-65MG is great), a good digi for the KK (set servo rate to ';high'
- Props must be perfectly balanced in both axis
- If you are taking off and landing on hard ground, protect the feet with something otherwise the carbon splits

I';ll get some vids up once I get a flight in daylight 

[Hands0n]

Yeah, thanks Revs, I look forward to that.  Thank you for the learned tips, they';re great.  I';ll have that printed out when i';m doing the assembly.

My motors, already bought, are SunnySky 980kv, they are small, don';t carry too much weight, and amazingly efficient. So i';m hopeful ::)

Re the KK2. I was totally negative about the KK2 flight controller after my dreadful experience with it coming to this hobby 20 months or so ago. It was dire!  But all that changed late this year when I picked up a cheap KK2.1.5 and popped Steveis 1.1.8S software on it. The stability and precision of the board with the new software is amazing.   It even has "Height Dampening" that implements a pseudo Altitude Hold in the KK2 without it having an actual barometer!! 

See here:
http://youtu.be/fDgBJsD6RIM

I';ll have to get me one of those analogue servos though - in anticipation I had picked up a digital servo (metal gear of course) so that would have been an interesting exercise!!

[Revs]

I had the KK2.1.5 in first. It was helpful in learning about PIDs in particular. I spent ages messing with the settings but never really got it so I was happy with it. I';ve since learned I was using old firmware (1.19) thinking it was the latest. But yes, the height damping works surprisingly well

I';ve got 1.18S1pro on the 250 and I still don';t think the auto-level is great, but I am still fine tuning it so I';ll give it more time.

[Hands0n]

Interesting that. I';m running 1.18S1 on Old Planky and it flies out of the box, no gains adjustments at all.  It could be a size thing, I';ve not mounted a KK2.1 on anything smaller, thats on the To Do list of experimentation ::)

[guest325]

Interesting that. I';m running 1.18S1 on Old Planky and it flies out of the box, no gains adjustments at all.  It could be a size thing, I';ve not mounted a KK2.1 on anything smaller, thats on the To Do list of experimentation ::)

I think you will find on a small model it will either feel a bit dead or like a mad squirrel (not quite sure which) 

[Revs]

It';s fine in acro but in SL yes it feels a bit dead. Slow to react and not very confidence inspiring.

[Hands0n]

It';s fine in acro but in SL yes it feels a bit dead. Slow to react and not very confidence inspiring.

Can';t you adjust that out using the SL gains?

[Revs]

Can';t you adjust that out using the SL gains?

Not easily it seems. If I put the SL gains up any more it gets twitchy. I can get it to move better with stick scaling, but then it';s nuts in arco. I';m trying to find a balance between SL settings, acro settings, stick scaling and expo on the tx.

The main issue at the moment is either a lack of time or a lack of sunlight. Should get more time over xmas.

[Hands0n]

Yeah, this time of year it all becomes about 98% theory  

[Revs]

I';ve just received that Tower Pro MG90S servo.. two days late. Shockingly poor quality servo. Not only is there a fair bit of backlash in the gears, but the output shaft has loads of slop in it in all three axis. You can grab it and move it up/down/left/right at least 0.5mm and pull it in/out even more.

I wouldn';t put this in a plane, never mind my tricopter.

EDIT: Just before I binned this servo I thought I';d see how easily the gears could be damaged. I put a horn on and on the first reasonably sharp twist the gears slipped.

AVOID!

[guest325]

These are my favourite for tricopters, but I have been trying 9gm mg';s lately on smaller models
http://www.hobbyking.co.uk/hobbyking/store/uh_viewItem.asp?idProduct=28416

[Revs]

These are my favourite for tricopters, but I have been trying 9gm mg';s lately on smaller models
http://www.hobbyking.co.uk/hobbyking/store/uh_viewItem.asp?idProduct=28416

Ah yes I';ve seen reviews of those and they';re quite well regarded. Is there much backlash in the gears?

This Hitec one, while expensive, it excellent. It';s has almost no backlash at all and is very smooth.

[guest325]

I';ve not had a bad one yet, used them in planes as well,  bit more oomph than a 9gm but not ridiculously bigger or heavier!

[Revs]

I';ll give ';em a try at some point. Cheers bud

[guest325]

I';ll give ';em a try at some point. Cheers bud

No problem,  JT uses them on his Tri';s - I pointed them out to him a while ago!

[Revs]

I';ve been skewed by Gav  cheers bud, can';t wait to try them.

[attachimg=1]

[Revs]

Thought I';d do a quick update to this thread

I dusted off the winter dust from the tri a few weeks back and had a fly. She flew fine, but something that has been niggling me with it; a lack of poke. Together these motors are rated at 69A/1.2kW and are only lifting 1.5kg inc. battery. So I had a look in the logs and it seems on max throttle it';s only drawing about 30A of the 69A it';s capable of.

Time for a change of props, so I bit the bullet and spent a rather large chunk of cash on these badboys..

[attachimg=1]

These should give her the extra lift I want while increasing efficiency. Not looking forward to retuning it as I';ve just done that, but hey ho, it';s godda be done.

If this weather holds up I might finally get some decent footage to post up soon

[Revs]

Well, bugger the yaw assy issue is back. I had a flight with the new props last night and the yaw assy worked loose again. I';ve decided I';m asking too much of this frame so I';m gonna retire it for now and maybe use it in a future project with smaller motors/props. The servo is also not coping with 14" blades and there isn';t room for a larger one. These new props are getting much more from the X3508S motors, though. It climbs at a hell of a rate now.

I';ve ordered a HMF-Y600 frame and I';m gonna transfer everything over. This frame looks way stronger, it has a more substantial yaw assy and a nice undercarriage which will allow for better camera mounting (ie, on a gimbal). Should be here tomorrow all being well.