AH-6Z Littlebird Night-Ops build project

[ClippedWing]

I guess if the 'Lead In' is off, the cyclic input is introducing a lag in an unwanted axis somewhere, and the FC not quite sure how to respond.

That would make sense i guess, but how are you going to change the phasing in angle while the FC just wants to correct it all the time? Sounds like a right head ache, and perfect for a bit of hammer time to me   :+

Touching back a few years, I think i cheated a little and use a STM or SMT (Not sure if its Swash Time Mix or Mixing timer lol) to sort out a Heli but it still never quiet got it 'Perfect'

Trying to work out phase lag, Meh rather you that me! This kind of stuff is on the limit of my aerodynamic knowledge  :-[

[Zeeflyboy]

Pretty simple, there is a slider that allows you to adjust phase angle electronically... At the moment it's at negative 40 degrees in order to give a mechanically 90 degree phased head.

My plan is simple, and more akin to your hammer method than any advanced aerodynamics calculation. I'm just going to increase the gains right up to the point that the oscillation just starts again. I will then change the phasing to negative 37 degrees and see... If the problem gets worse, then I will try negative 43, if the problem reduces but is still there I'll try negative 35 etc. Once the oscillation goes away (if it does lol) I'll increase the gains until it comes back and continue fine tuning phase angle. If the oscillation reverse direction, it probably suggests that I've gone past the correct setting and need to dial it back a bit. Once I get the gains high enough that it starts bouncing on axis but not oscillating, I'll know I've found my phasing angle and then dial the gains down enough to stop the bounce.... Then I'll try atti mode again and see if my problems persist.

That's my theory anyway. I'm just glad I built this training gear because intentional forcing the heli to oscillate would get my knickers in a twist otherwise.

[wgt40w]

So, all in all it looks like you are going to build a mark II chassis and mechanics + mark III on the electronics side !

You sure are a glutton for punishment - but at least you will be giving your 'followers' much more to enjoy reading about and being even more in awe of your ability to learn and apply building / machining and electronic skills !!

[Zeeflyboy]

Well I'm optimistic that with the help of Chris Bergen I may be able to sort my Ace One issues out... if that is the case then I will (at least for now) continue to use it.

The new mechanics are really just a "because I can" thing... back when I started this thing I didn't have the same knowledge or skills that I do now and settled just for something that is essentially an (admittedly fairly extensively) modified t-rex 600e.

Now with the equipment I have, and the design skills I'm still learning, I feel I can aim a little higher and design something from scratch that will be targeted specifically at what I need. The main points are high gear ratios achieved through compound gearing to better suit low head speeds, high tail to main rotor gear ratio to enable better tail authority at low head speeds, and finally reduced vibration and noise by running a dual main belt drive from motor to main shaft.

Like I said it's really just something I'm going to develop on the side - once the Ace One (or pixhawk if necessary) is running as I would like, I can install the present mechanics in the fuselage without waiting for the new one.

[Zeeflyboy]

Been doing a little design work on the mkII mechs... even modelled a Kontronik Pyro 650 to help me figure out placement of everything. The Pyro will probably replace the Scorpion, it puts out a max continuous power of 3kW (about 4 horse power) which is way more than this needs but weighs 290g vs the scorpion's 460g.... that's a huge difference. Supposed to be an extremely efficient motor too, and considering I shed 170g (or optionally gain 170g of battery capacity) that should help flight times.

The catch is that the pyro 650 only comes in a minimum of 620kv which is waaaaay too high, so I have sent an email to Gmauro (the brushless motor rewinding king) to get a price on one rewound to around 420kv.

anyway, a quick look at the mechanics starting to develop (ignore the frame at the front, it's still based on the old frame's dimensions:

[Zeeflyboy]

Had a little time at home to start the first part of the mkII mechanics - the main gear. On the current gear setup I have an aluminium hub housing a single radial bearing and a one way roller bearing, which is then screwed into the gear. Because I'm making this gear out of aluminium I have no need for a separate hub and have designed the gear as a single piece, plus added a second radial on the bottom.

the design:






and I'm cheating a bit here, using a pre-made pulley:






So I pre-machined a jig to centre the piece in X and Y, plus level the bed at the same time.




Then I glued the part in (using a glue that can be dissolved in acetone)




and after machining, I cut down the surrounding mdf and soaked the part in acetone for a couple of minutes, the glue dissolved and the part was released:





So that's only one side obviously but I ran out of time and had to come back to Geneva. The next job will be flipping it over and machining the bottom side.

Now that i'm back though I'll be doing those adjustments to the ace one and seeing if I can get this thing flying a bit better.

[Zeeflyboy]

A bit more design work... it's going to be a fixed belt drive system - distances between pulleys have been specced to the most suitable length of belt available, and any belt adjustment will be automatic in the form of sprung tensioners.

Tomorrow i'm on standby for work - if they don't call then hopefully i'll get to do a couple more test flights to see if I can nail down the Ace One.

[ClippedWing]

I hope you get a few flights in today and are able to adjust it a bit more to your liking!

I love your simple ways of getting over engineering projects like gluing the block to the board!

Fingers crossed it goes well!

[Zeeflyboy]

not today, tomorrow - worked today :( hopefully they leave me alone tomorrow and I can play lol.

So just one more pic cos I think it looks cool - a close up of the drive train now that i've mocked up some pulleys.




So it's a compound drive... 14T on motor driving a 40T for a drive ratio of 2.86, then 14T driving a 60T with a drive ratio of 4.28... so unless my compound gear maths is wrong that gives a total drive ratio of 12.24 in a space less than 60mm wide. Assuming a 420kv motor on 12S, that's a rough max HS of 1523rpm... at 80% for a good governor that's 1218 or so. Should be pretty much in the sweet spot I'm aiming for.

The tail drive will be 60T to 12T for a simple 5:1 ratio (slightly higher than the current 4.5:1 in order to give more tail authority).

Assuming I manage to build it all as planned, should be a nice smooth power system too with the belts reducing gear mesh noise and vibration.

[wgt40w]

Whose going to join me in betting that within 6 months - after successfully testing the Mk II mecs, Zeeflyboy will be considering the Mk III mecs - all shaft driven !!!

PS See you've (Zeeflyboy) just bought a cnc from Gordon C - seem to remember you buying your first one whilst in Switzerland - does this mean you will now have 2 ?? or has the first worn out ??

[Zeeflyboy]

MkIII will be turbine driven :) lol

Re the CNCs, think you've got the timelines a bit messed up. I bought my first CNC from Gordon here almost 2yrs ago... I need to get around to selling it actually.

The CNC I am using now is a considerably more powerful beast (X6-1500GT) with much bigger stepper motors, ballscrews etc and a larger working area. To give you an idea of the difference, the spindle for the old CNC was a 100W proxxon router, on the current CNC it's a 1.5kw liquid cooled brushless spindle...

[wgt40w]

MkIII will be turbine driven :) lol

Re the CNCs, think you've got the timelines a bit messed up. I bought my first CNC from Gordon here almost 2yrs ago... I need to get around to selling it actually.

The CNC I am using now is a considerably more powerful beast (X6-1500GT) with much bigger stepper motors, ballscrews etc and a larger working area. To give you an idea of the difference, the spindle for the old CNC was a 100W proxxon router, on the current CNC it's a 1.5kw liquid cooled brushless spindle...

Glad to might think about a Mk III - means this very enjoyable and interesting Thread will last longer !!!

CNC - first one 2yrs ago - gosh, doesn't time fly - it was a result of reading about yours that I eventually bought a similar one.

[Zeeflyboy]

Glad to have inspired you, just don't send me the bill lol.

Just been out for some more test flights, not a whole lot of progress tbh. In manual mode the -40 degrees I already had set for phasing seemed the most correct position, I tried a range from around -37 to -43.

It's flying ok in manual mode, still not as locked in as I'm used to with traditional FBL controllers but manageable. In atti there is still this persistent left drift, and in GPS it holds position fairly well but will still fly off to the left at an angle if you push straight forwards.

If I'm honest, it could be flown like this. I could go into the taranis and add a right trim input activated with the atti and mix in a little right stick with forward stick in GPS atti to get it flying straight... After about 6 test flights with very little signs of improvement I am tempted to do that just to get it flying.

I am however getting ever closer to throwing in the towel and getting a pixhawk for this thing. The only thing holding me back is that nagging doubt about reliability and possible susceptibility to vibration.

[wgt40w]

With the tremendous effort, skill and detail you have so far put into this project - to be "flying ok in manual mode, still not as locked in as I'm used to" is disappointing.

I would really hate it if you assembled the detailed body back on and had a major accident.

Please, please, persevere while the heli is in chassis form to sort out the or a controller to give you the stable level of flying Juan deserves.

[Zeeflyboy]

I'm pretty much committed to the pixhawk now, I've been doing a lot of research and talking to one of the chief developers of the traditional heli side.

It's really getting incredibly advanced now - they have even recently implemented extended kalman filters (basically an improved method of filtering out bad data that could be introduced by sensor failure, vibration etc) and even use the dual IMUs and have support for dual GPS etc.

Rob showed me this video of a fully autonomous flight on the latest firmware... it just amazes me that this thing is doing this completely without human intervention... especially those nice curving spline nav turns. It's worth noting that the Ace One GPS speed is limited to around 10m/s - this flies at around 30m/s in this video. I am also loving the fact that it can RTH and land, the Ace will just RTH.

https://www.youtube.com/watch?v=kQj50ykhEnQ


Also a bit more work on the mechanics and it's starting to flesh out a bit:






At the moment it's slightly simplified. There will be an idler pulley to increase tooth engagement on the smaller tail pulley and tun a couple of tensioners too once I've got it all figured out.