Tradheli from a beginners POV

Blog
,
1

b50dcc0c935c12e465feb88be911253e411bd07e.jpg1920×1080 640 KB

I’ve been flying fixed wing mapping planes for the last couple of years, but a few months back, Rob Lefebvre’s Protcyon project caught my eye and I found the concept intriguing.
But…Helicopters are expensive and complicated right?. Turns out neither is really true.
Sure, it isn’t as cheap as a throwing an APM on a foamplane, but you can still get in the air for well under $1000.
Reading some more and asking a few questions, I started putting a shopping list together.
Initial plan was a 500-sized heli as they can actually carry stuff and parts are plentiful and cheap. I was going to go with a Trex clone, but at one of the local drone races, there was a guy doing a demo with a Protos and despite being labelled as a high-performance brand, it was still reasonably priced. Better still, the Protos is also a lot lighter than it’s contemporaries.
I still didn’t have a $1000 lying around for a potential risky project, so I put the idea on hold, but still kept browsing the classifieds on RCgroups and Helifreaks and then a deal too tempting popped up and I was suddenly the owner of a 2nd hand Protos 500.

First task was making a tray for the pixhawk and the anti-vibration mount.

mount1.jpg3264×2448 1.44 MB

It’s not pretty but it will do for now.

Then adding the usual accessories; Power module, GPS, telemetry and a Brotronic Micro-UHF for control.

Bench test worked beautifully, but then when I got it all mounted, my servos were jittering like mad. Googling (and calling a friend!) suggested that it was the 433mhz telemetry from the micro-UHF, but disabling that and moving the RX & it’s antenna away made very little improvement.
I then realized that my 915mhz module was a 500mw module and moving that away immediately solved the issue.

Version1.jpg3000×2051 509 KB

Everything is now set up and working. I’ll go through the swash plate leveling and pitch setup again, but it essentially ready for it’s maiden.

Future plans involve using a Pixracer which I should be able to mount in the frame tray. Then a couple of custom carbon fiber landing skids, which will include a mount for a mapping camera and a 3 axis Storm32 gimbal.
A custom canopy is also planned but the turnover for that will probably be slow.

I’ll update this blog as I go along, so stay tuned!<img

6 Likes
2

Looking good!
I was wondering if you could move the first picture up near the top? I think that will make it appear on the main ardupilot.org page even before people enter the blog. Thanks!

3

Nice and simple! :+1:

I also bought a second hand Protos 500 about a year ago to do the same thing but never got around to finishing it! Other projects got in the way.

Yes, please keep us update on your progress.

Thanks.

4

Done! I’ll swap it with a more attractive one later.

I should also add that I still have a bit of servo jittering now, so I will probably swap that 500mw module out with a standard 100mw from one of my planes. (Where it’s much easier to isolate it from the servos). It’s also unlikely that this one will fly far enough away for me to ever need 500mw/

5

That’s great. I’ve set myself the task of building the cheapest Heli I could. Rob L. suggested 500 size as the minimum so I bought a cheap clone one. My thinking is I’m going to crater it fairly early on so may as well be a cheap crash. I’m up to configuring my Taranis controls and am a bit stuck - need to do more reading. But I’ve got an ace Heli pilot at my RC Club so he will maiden it just with a gyro to make sure all the mechanics are fine and I will then have a couple of flies and if it isn’t a smoking crater by then I’ll fit an autopilot between the skids and away we go!

Thanks, Grant.

6

I have a Chaos 600 that I converted to Electric. It still has a flybar. Will the firmware handle helicopters that use fly bars? Currently having a blast using the airplane firmware on my Sig 4 star 20.

7

You might find out that the cheapest helicopters are actually the most expensive ones. The bad quality of shafts and bearings of the clones forces you to buy lots of replacement parts of the ‘original’. This adds up !

8

With my Heli flying skills I’m expecting to crater it hard enough that it would be a full replacement. If you do something - do it properly I say :slight_smile:

Thanks, Grant.

9

O.k. I understand, please make a video so all have fun :slight_smile:. If you need help use Tradheli instead of a FBL-unit and play with D parameters in Extended Tuning :grinning:

10

Sounds great, Grant.

More of us trying trad heli the better!:thumbsup:

Cheers, Steve

11

So, it took a while to get the weather and spare time to co-operate, but it finally worked out today.
Took the first few spool up, I couldn’t get the tail to hold at all and first I realized that I had the tail-servo direction set wrong!. No harm done.
I got it up a few inches, but it was still struggling to hold the tail (but at least in the right direction) and it took a fair bit of faffing, before I looked down on my radio and saw that it had a huge handful of rudder trim on it.
The profile was initially copied over from one of my planes and I hadn’t reset the trim! Doh!

So, up I went again…and it flew!. Now it was a little more windy that I would have liked and while I was relatively shielded from, it meant that it was a bit swirly, so it was real hard to judge the tuning,
I fiddled a bit with the roll & pitch P and it got a little shaky, so dialed it down a tad, but it still felt quite drunk…but controllable!.
This was all done in stabilize and I decided to brave it and switch it to loiter…and once I had the throttle set at hover, I could take my hands off the stick, which put a real big smile on my face!.
I repeated this a couple of times and on the last one, I could see it struggling a bit to hold and was starting to chase itself, so decided to call it a day, so I could go home and look at the logs.
Looking at it, I still need to be more aggressive with the P (or FF?) but at least I now know that it flies!.
It pulls just under 15A in a hover, with a head speed of ~2400, which should ~15 minutes with the 5000mAh 4S.
Once I get it flying better, I’ll start playing with lowering the headspeed and adding asymmetrical blades.
I also have some Multistar 10A lipos, which are only about 200grams heavier than my 4S, so it looks like 20mins with a mapping camera should be doable.

The only real issue is that my diy 915mhz half-wave circle antenna is a bit too directional and I struggled initially to connect to the heli, when the heli was below the GCS.

Unfortunately, my mobius had discharged itself since last time I (barely) used it, so I have no video and I still didn’t trust it enough to put the radio down and get my phone out :slight_smile:

With a bit of luck, I’ll be able to get out again tomorrow after work. The real advantage with this, compared to my planes, is that I can test it locally, rather than having to head out to the sticks.

12

As answer to this question. Yes, the firmware will fly a flybar with no problems. They are actually incredibly stable with flybar. I consider a flybar helicopter of 600 or 700 size class, flown by ArduPilot, to be one of the best VTOL camera platforms there is.

13

I am flying a Trex 500 with a Multistar 10A 4S, using a 6S motor and gear train. 30 minutes is no problem at 2,000 rpm headspeed. I built a taller and wider landing gear for it and slung the big 10A underneath with the Pixhawk in back of it. Mine pulls about 19 amps in flight at 6lbs takeoff weight. One thing I’ve found with those Multistar 10A’s is that if draw them down to ~3.8vpc at rest, after the battery has cooled and recovered, it takes roughly 17,000 mAh to recharge it, according to my charger. Some people would say that I’m over-discharging it, but I got one that’s got over 100 cycles on it flown that way, and the cells are still even and it delivers as much power as it did when new.

They’re not great for getting maximum power, as the voltage does tend to “sag” a bit more than a 35 or 40C battery under load. But they work really well for a long flight time 500 heli in normal flight where you might only get up to 22-23 amps in a turn with the collective ramped up a little.

This is how I mounted that big battery to my Trex 500. With the battery mass slightly forward of the mainshaft centerline the heli balances out pefect. That battery weighs 1.8lbs.

https://goo.gl/photos/YNZdU492HqXTK4Zb6

14

Markus, I’m using rate controller D-gain in pitch and roll with my 500DFC, and it’s every bit as stable as my flybar heli’s with 3.4.5. I kinda think that 90% of the stability problems in trad heli are due to possible problems with mechanical setup, and not necessarily with the software.

One parameter to watch out for, though, and this is not well documented, is ACCEL_Z_P. It comes in trad heli set to the multi-rotor default of 0.5. That will make the collective cycle rapidly on most heli’s, and switching to Alt Hold before you get your mid-stick collective set right in Stabilize will likely end with getting your heli slammed into the ground if you’re only at 10-15 feet altitude. Drop that to 0.20-0.25 for your initial flight and then tune it from there. I’ve found that 0.25-0.30 is where my 500-700 size heli’s ended up without getting collective “bounce”. And that’s with my servo throws at 13mm. If you use 16mm your heli is likely to turn into a Wild Thang set at 0.5, and you will gain instant respect for much more responsive a helicopter is than a multirotor :grin:

15

Yeah, I saw that picture in the ‘parameter’ thread and I’ll probably copy some of your landing gear design.
I want to make it out of carbon fiber, but may mock up a aluminum version first.
The plan is to move to a Pixracer or MicroPix, which should fit on the rear ‘tray’ of the heli, leaving the bottom of the frame/landing gear for battery and camera mounts.

16

The stock landing gear on a Trex 500 (or any of them for that matter) is a joke. It’s only good for landing on a flat hard surface. Land in the grass with it and you’re almost certain to be mowing grass with the tail rotor. The first time I flew it with the stock gear on it I flared and set the heli down in the grass at the same time it came to a stop. Zing - there went a tooth on the tail drive gear.

So I decided the first thing it’s going to get is a real landing gear with some ground clearance, that I can at least touch the rear of skids down in a flare without banging the tail rotor in the dirt.

If you copy it, note I did not use the stock landing gear mounts either, on the frame. I made cradles that the frames sits in, and the frame bolts to the cradles with stainless steel bolts on the side. The frame in my Trex is also CNC milled aluminum, and not carbon fibber. I built it from ground up as a test heli, using a Trex 500 as a pattern (along with some Trex 500 parts) to be able to crash it, pull it out of the bottom of the smoking crater, dust it off, reflect briefly on how stupid of a move that just was, and go flying again. :grin:

If your heli was chasing itself in Loiter, which is not uncommon even with a good Stabilize/Alt Hold tune, you’re probably right - it’s needs more P gain. Not every one is the same, unfortunately. There can be differences in servos, what horn you use on the servo, or even difference in the head from one make to another. But at least you got it to fly, and that’s a major first step!

Did you try any cyclic in Loiter? To see if the heli responds, or if it’s really sluggish?

17

It was fairly sluggish. Weather forecast for the next week + is pretty miserable, so it’ll be awhile before I can get it out again. Guess I can start on that landing gear :slight_smile:

18

It will be interesting to see your mockup with aluminum, and how you transfer that to carbon fiber. CF is more flexible and doesn’t have the rigidity and strength in the cross section of the .120" flat stock I used for the main gear legs. The entire rest of the gear is .060 x 1/2, except for the cradles on the frame.

19

Carbon fiber is stronger and more rigid than aluminum, depending on how it’s laid up/constructed.

I did a quick test a few weeks back, which was also about 1.5mm (0.06") thick and yes, it was clearly too flexibly, but that was done as a flat lay-up with all the layers in the same direction.
So, either putting alternate layers on the 45 degree bias or turning it into a more ‘3d’ like shape with a core material should make it plenty rigid, saving an awful lot of weight.
The downside is obviously turn-around time, which means I really want to get the design right the first time around :).

20

That will be really interesting to see your gear. I’ve never worked that much with carbon fiber, and would like to learn more about it.

Not a lot of people do long flight time low-headspeed 500’s because the common misconception is that it doesn’t work all that well and you need at least 2,300 rpm to get a 500 to fly decent. You really don’t. Mine will fly at 1,800 just fine. Not going to do sport aerobatics with it at that headspeed at its 3,000g takeoff weight, but it flies really nice for scale flight. I usually run it around 2,000 rpm and it’s got decent performance at that headspeed with symmetric blades (stock 420’s) and hovers on about 6 degrees of collective.

I find it hard to wrap my head around why most folks think flying a 500 heli at 2,800rpm that can pull 10g maneuvers is the ultimate. So this is a really nice project you got going here! Are you going to put a bigger pinion on it to slow the head down?