AH-6Z Littlebird Night-Ops build project

[Coyote]

Great news mate, seems to be holding up fine now

[Zeeflyboy]

Cheers guys, was really nice to have a successful flight (well, hover mostly lol) with it.

I think it's going to need a fair bit of tweaking to get it flying to how I want, but I'm just happy the Ace One isn't broken!

Next big step will be testing auto modes, but once that's done the opportunity for big setup related crashes should mostly be behind me! I want to shake it down with a few more manual flights and tuning before I move on to that stage.

And of course, I better get back to work on the fuselage.

[ClippedWing]

Take you time on the tuning! Nail it right in!
When I say take you time, try and get the tuning done within 1 year lol!

How is the Ace One For turning? Is it a right faff?

[Zeeflyboy]

Erm, not sure yet really but it doesn't look any worse than a typical heli FBL controller.

In manual mode you basically have 4 things to balance out: gains (fore/aft, left/right, tail) direct input, cyclic rate, flybar weight.

Then in auto modes you have the separate gains, stick delay, velocity tracking.

I have the added complication of tweaking the e-phasing to minimise aileron/elevator interaction too... I think I should probably start with optimising that.

Compared to some heli systems though (such as my sk720 for example) there are very few parameters overall to mess with - Hopefully it should be relatively easy to dial in to a satisfactory level.

[ClippedWing]

We talking my satisfactory level or your satisfactory level lol!

Im looking forward too seeing the mech and canopy on now! 99% of people would of given up by now, how do you keep going?

Look forward to more news soon

[wgt40w]

I bet Juan can't wait to get in it now!!

[Zeeflyboy]

While my batteries are charging for another couple of test flights this afternoon, I've just been doing some work on the project i've now dubbed AH-6 Xtreme for my work folder ;)

Couple of improvements over the current frame so far:

- Tail servo moved in from the tail boom to the frame to help CG.
- Middle bearing block accepts a large angular contact bearing (accepts high radial and thrust loads in one).
- Improved servo geometry, also with no linkages or levers - all direct to swash.
- Ace One IMU mount moved inwards closer to the CG, better for accelerometer accuracy and also helps with overall CG.

I am planning to use a dual belt drive system with compound gearing. I should be able to achieve around 14:1 on the motor to main rotor and then 5:1 on main rotor to tail - this will give more tail authority even with the low main rotor RPM... an improvement over the current 12.5:1 and 4.5:1 of the current setup. I keep reading how smooth the dual belt setups are too so should hopefully keep vibrations low.

I can't design the drive system until I have pulleys and belts in hand though, so now work stops for a while :(












good view of beefcake 10x26x8mm angular bearing:

[Zeeflyboy]

So, just got back from two more test flights, second one without the training gear ;)

I've got the cyclic behaving fairly nicely in manual, but the tail I can't quite get right, either too loose or too bouncy. I think I have a mechanical gain issue there so I need to get/modify a shorter servo arm there  methinks.

I had a very quick flick into Atti a couple of times but it quickly started a rapid drift backwards and left - will need to recheck my swash is level.

However, another 20 minutes of airtime and it's still in one piece!

[ClippedWing]

I'm quite happy to see the AhZ hoovering in one place for 20min, anyone else lol?

Videos videos videos!

[Zeeflyboy]

doesn't make for very exciting videos! wasn't just hovering though, was some very careful moving around too ;)

Hopefully going to fly tomorrow too after some tweaks tonight, so we'll see if anyone is around to grab some more video.

[Zeeflyboy]

hmmmm... mixed bag.


Flew another two flights, I have manual mode flying ok-ish, and the tail is better after shortening the servo arm - I think it could stand to go another 2mm shorter or so to get it just perfect. In order to have it not constantly drift to the left and back I had to trim my swash out of level.

I also played around with Atti and GPS atti, including switching between manual and atti/gps atti in flight.

That's where the mixed bag comes in - it didn't crash (yay!) and switch over was smooth between modes. The issue is that upon selecting Atti the helicopter starts a strong drift left and slightly backwards, even with the swash trimmed to fly level in manual mode. humph.

So anyway, on to testing GPS atti and it holds pretty decently in the hover after an initial bump to the left due to having to pass through atti to get there. Weird one though - push forwards and the heli drives about 20-30 degrees left of track rather than straight ahead. And yes I calibrated the compass, twice. The DJI ace one wiki says that if it tracks to the side, then you should rotate the compass. Thing is that's a big offset and where I am is near to 0 degrees declination. I've never had to twist my Naza pucks.

Sigh... I really feel like a product this expensive should be a bit more straight forward. I need to do some reading up on my problems and see if I can find someone experienced with the Ace One.

[ClippedWing]

That Ace One will be the end of you!

Did you get the ace one looked at after your first crash? I cant remember? Turning pucks and compass` ofset manualy sounds bizzar to me, and wouldnt sit right with me either!

The whole point of the system is to make your flights stable, not drift when it wants too!

Im waving a red flag on that Ace One unit you have!

[Zeeflyboy]

Yeah, it's a bit frustrating.

The more I read about the pixhawk and the direction it's starting to take now that they are beginning to leave the APM hardware behind (e.g. enhanced kalman filters is a big one for system reliability and rejection of bad data) the more interested I get.

The new mechanics I showed above will be wide enough to accommodate a pixhawk, so I've got my bases covered should I wish to switch. The big issue is money now lol, having bought a house and moving back to the UK with a pay cut the toy funds aren't quite what they used to be.

[ClippedWing]

Oh boo hoo, you already have the best toys!  :laugh: :laugh: :laugh:

Is there not somone local to you with ACE One expert knowledge? (Dont look at me im a hammer and brute force kinda guy)
I would leave the Pixa for a year, let them work the bugs!

Again, for the money that Ace One should be more than enough to sort the AhZ out! Might even be worth a send to the DJI folk (not that i heard there very good at helping customers) Does the unit do data logging?

[Zeeflyboy]

I've never had a reply of any actually use from DJI, and that's when I get a reply at all.

There is a DJI expert (Chris Bergen) I've asked for advice, and he has kindly suggested the problem may lie in my swash timing. It's currently set at negative 40 degrees to bring the mechanical phasing to 90 degrees but he suggests this, while the best place to start, may need adjusting in flight testing.

It would explain a couple of anomalies... That circular oscillation I mentioned on the very first test when the gains were a bit higher. Normally if a gain is too high in an axis on a heli I would expect a bounce in that axis. If the phasing is off to the extent that the FBL unit's correction in aileron is actually also introducing elevator then you may end up with this rolling correction resulting in the rotating oscillation I saw.

So, next job will be resetting the swash to level and try playing with the phasing angle.