Six-rotor helicoper with Arduino

Go To Last Post
42 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi guys,

take a look at this:

http://www.crunchgear.com/2010/0...

Pretty cool...

*** Eivind, webmaster ***

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Holy mackerel - that thing has a LOT of power to take-off like that - wonder how long the batteries actually last? 4S LiPo is 14.8V and with a 3300mAh capacity I'm guessing they drop at least 10C so that would be 33A for 6 minutes of flight. 33A * 14.8V is 488W. So almost half a kilowatt - that's quite a beast! (quite a potential to kill if it flew out of control too I guess)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The helicopter is brilliant!

I had never done the sums for battery energy. On this basis a 2500mAh NiMh D cell will do 30W for six minutes.

Now I understand why things can get hot!

I also see how these electric models can fly. I had always assumed that electric motors would never compare to an internal combustion model motor. But I suppose that model fuel would always win for long endurance flights.

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

With Brushless you are at least looking at probably 75-80% motor efficiencies (brushed can do down around the 50% 's!) so of the 488W maybe only 15-20% will dissipate as heat but that's still something like 85W going to heat generation (luckily the motors are air-cooled of course!).

In all my R/C flying I've never (apart from a half hour helicopter flying lesson with one of the UK's top RC heli stunt pilots) flown I/C. When I first started it was just on the transition from NiCd to NiMh (more power density but higher internal resistance) and I/C flyers used to poo-poo the electric flyers as a lower social species, rightly (at the time) suggesting that electric systems couldn't match I/C. This wasn't helped by the fact that brushless ESCs and motors were so astronomically expensive that we were all pretty much limited to brushed motors and ESCs (and dissapating half the battery capacity as heat didn't help the electric flyer's cause much).

Then electronics got cheaper, mobile phones and laptops pushed the development of LiIon and later LiPo cells and slowly those cells moved from being able to dump 1C to 5C then 10C, now 20C and headed even further. At 20C you only get 3 minutes (but 20 times the rated current) but you can have a lot of fun.

multi-MOSFET brushless ESCs came down to the previous price levels of brushed ESCs and (partly because of a lot of us rewinding CD drive motors with N45 neodyium magnets and huge amounts of copper) the cost of brushless was foced down (previously they'd been hand constructed works of art as it's far easier to wind a brushed commutator on a machine but new production techniques made machine wound brushless possible). So now we had 80%+ efficiency power plants to work with our extremely high energy density battery cells.

Personally I tend to fly 100g..200g aircraft using 50W..100W of power, usually around 5C..10C giving me 12..6 minutes of flight. The idea of 500W aircraft would scare the living bejesus out of me! (a 1.5Kg, 200W helicopter once had a very unnerving effect on me!)

Bottom line is that these days electric can easily give I/C flyers a run for their money and if anyone is now in a position to gloat it's the electric flyers.

Eivinds video is just more grist to the mill for the advocates of electric flight.

(wonder when we'll see this translate to full size aircraft! ;-) )

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Eivind,

Thanks for posting the link, that is an INCREDIBLE video. Stable, fast, powerful, controlable...

Very well done project!

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That is freaking cool! Thanks Elvind!

Stu

Engineering seems to boil down to: Cheap. Fast. Good. Choose two. Sometimes choose only one.

Newbie? Be sure to read the thread Newbie? Start here!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Do any of you guys know of a model sized gas engine > electric generator combo? You'd think that this would allow much longer flight where one gas engine provided the electricity for the six electric motors. I know that some of the military robots with lots of electric motors on them, have gas generators for power, but I haven't seen anything small enough for this application.

Smiley

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Joe,

It's all about power to weight lifting ratios and energy density of the "fuel". I forget what the equivalent actually is (you'd find it in a search of rcgroups.com) but the power density of nitro versus the highest density LiPo is now pretty close. So then it's a question of motor/mounting weight and I think for a given power output electric now scores slightly higher than I/C on this.

Unlike a robot an aircraft has to lift it's entire weight off the ground. A thrust to weight ratio as low as 33% is enough to get a fairly lethargic plane off the ground. Sports planes need about 80%..120%. To hover a plane on the prop you need about 150% (you might think 100% would be enough but it isn't) and for really serious 3D you need about 200%. Unless Nitro was hugely more power dense than LiPo you couldn't justify carrying the weight of a an I/C->electro generator. In fact you might as well stick a prop on the I/C in this case.

Cliff

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why does he need six rotors? Isnt 3 enough? (defines a planar surface?). It cant yaw while hovering, evidently... need to pitch fwd, get up some fwd speed, then bank?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd say you need an even number of motors to get the torque evened out. Then the choices are 4, 6, ... and smaller motors are cheaper. That makes it 6.

Edit: And of course, it can yaw. There's just no point in doing so (for demonstration), because the whole thing can travel in any direction anyway. Yawing is done by redistributing torque while keeping the lift at a constant level.

I tend to post off-topic replies when I've noticed some interesting detail.
Feel free to stop me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Reminds of a scene from the 80's movie Runaway with Tom Selleck.
There was a scene in the movie where the police used a thin small helicopter
with a camera kind of like this model to investigate a building to locate
the "bad guys".
Theirs didn't look to plausible as it had only 1 rotor.

--- bill

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You need to explain to me how he can change heading while hovering. You say take one motor to 100% and bring the other 5 down 20%? That will just pitch up on the full speed rotor side, not yaw?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

take a quadrocopter: You have 4 motors, I'll call them (N)orth, (E)ast, (S)outh and (W)est.

N and S turn clockwise;
W and E turn counter-clockwise.

Hovering: All motors turn at 80% speed, all "torqued out".

Turning: Now set N and S to 85% (more counter-clockwise torque), W and E to 75% (less clockwise torque). The copter will turn counter-clockwise.

I tend to post off-topic replies when I've noticed some interesting detail.
Feel free to stop me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

buffi wrote:
I'd say you need an even number of motors to get the torque evened out. Then the choices are 4, 6, ... and smaller motors are cheaper. That makes it 6.

Tri-copters yaw (or prevent yaw) by having the rear motor tiltable. This video shows it pretty well:

http://vimeo.com/5965644

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Grocious me .... that thing takes off as if it's a rocket ! Very impressive. And it shows what can be done with an Arduino :)

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

wow, it takes real confidence in your work to fly that thing in such a small room. I'm guessing it wasn't the first test ;)

Looking through the mikrocopter site it looks like it uses a custom board with a mega644, don't know where CruchGear article gets Arduino from. Has Arduino replaced pic as the layman's term for microcontroller?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Whoops. Never mind.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

smileymicros wrote:
Do any of you guys know of a model sized gas engine > electric generator combo? You'd think that this would allow much longer flight where one gas engine provided the electricity for the six electric motors. I know that some of the military robots with lots of electric motors on them, have gas generators for power, but I haven't seen anything small enough for this application.

Smiley

Joe, it is now 14+ years ago, but during my involvement with the development of the 3 metre wingspan Aerosonde, we used a 20cc heavily modified Enya model aircraft engine driving a pusher propeller with a takeoff to a small electric generator that charged our bank of NiCds for the servos, GPS and control electronics. We used 100LL aviation fuel because of its very high energy density.

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, I have to defer to Cliff because I don't know a lot about this. I'm surprised that the energy density of the Li-poly batteries is better than a theoretical IC motor/generator combo, at least after a certain size where the weight of one motor/generator becomes low compared to the fuel. All the cases I know about are military ground or sea vessels, not aircraft so I guess if my case was valid, somebody would be doing it.

Smiley

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, we were able to fly our 15 kg plane for 40 hours at 75 kph on 4 litres of fuel. Do you think 4 kgs of Li-poly batteries could sustain that workload? I really have my doubts.

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The Aerosonde does look impressive, given that you didn't build only one (I've looked that up on wikipedia). At what power consumption were you able to fly that long?

We're building a 5.4m fixed wing UAV here which will use LiFe batteries and a 2.5kW brushless motor. I'd like to make some comparisons just out of curiosity.

I tend to post off-topic replies when I've noticed some interesting detail.
Feel free to stop me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ross,

I think you are right about that but do you know the HorsePower figure for the motor/prop combo? It should be able to convert that to Watts (1 horsepower = 745.699872 watts) for an equivalent electric power to turn the same prop at the same rate then work back to P=IV where V dictates the number of LiPo cells (3.7V nominal) and the capacity (mAh) and discharge rate possible dictate I

To turn this on it's head. Say you have a mobile with a 50g LiPo battery that can run it for 20 hours, what weight of oil based fuel would be required to power it for the same length of time? (so gone are the days of mobiles with aerials, but now they have exhaust pipes!)

Cliff

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The thread title and info is incorrect.
It's a Mikrokopter, and it does NOT use an Arduino, the Flight Controller is their own, with a mega644.

Funny to see this now, I'm actually building a craft similar to that right now. It does not need much power at all, I expect about 15-20 minutes on a 1800 mAh 3-cell Lipo (about 19 Wh). That's with payload of a movable video camera and a 5.8GHz video downlink. Telemetry goes either on one of the download audio channels or on a 868 MHz RF link. The latter is two-way and allow updates to nav points, parameters e.t.c. in flight.
The Flight and Nav controller is my own design with all on one 50x50 board. Playing with some brushless motor controller also.
Much fun.

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It was around 800 watts give or take a margin for 14 years of "memory rust" ...

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd love to build one of these but looking at the site's shop page, it all cost WAY too much money!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

I expect about 15-20 minutes on a 1800 mAh 3-cell Lipo

20 minutes is 3C so you are planning to pull 5.4A from the cells. 3S is 3 * 3.7V = 11.1V. So that would be 59.94W. Usually that's enough to lift about 50..100 grammes in a hovering device, how much does the camera/mount weigh?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ross, from the fuel flow I calculated the following:

rho = 750 kg /m³,
V = 0.004 m³,
H = 43730 kj/kg,
t = 144000s
P = H rho V / t = 911 W.
Expecting a COP of 25 % (I have NO idea if that is realistic, it's probably much less).
P(Shaft) = 0.25 P =~ 230 W.

Now how heavy is a 230 W (Shaft power) LiPo drive train that can operate just as long? I can't answer that now because I don't know how much energy can be taken out of the LiPos.

I tend to post off-topic replies when I've noticed some interesting detail.
Feel free to stop me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

I don't know how much energy can be taken out of the LiPos.

As I say - cuurent design can drop up to about 20C. In other words say you combine cells in parallel (to increase voltage) and also in series (to increase current sourcing) then you might have something like a 3S4P pack of 2000mAh cells. The 3S means "3 in series" and because one cell is 3.7V optimal this would therefore be 11.1V. The 4P means 4 lots of this in paralle (so 12 cells in all) which effectively bumps the 2000mAh per cell up to 8000mAh for the pack. Now a 8000mAh pack could source 8A for 1 hour and this rate of current draw is known in the trade as 1C. It could source 16A for 30 minutes (2C), 32A for 15 minutes (4C), 40A for 12 minutes (5C) but for real performance LiPo tend to be used in the realms of 10C..20C. At 10C it's supplying 80A for 1/10th of an hour (6 minutes). At 20C it's 160A for 1/20th of an hour (3 minutes). Obviously P=IV so, say you designed for 10C usages then that would be 80A and 11.1V so this pack could deliver 888W for 6 minutes.

Here's some cell weights:

http://www.electrifly.com/batter...

So they have 3S 2200mAh cels there (for example) that weigh 179g. If my example above used these (rather than the theoretical 2000mAh I stated) they'd be 4P so 4 * 179 = 716g but 4 * 2.2A = 8.8A. While the cells say they can do up to 25C I think 6 minutes (10C) is more practical so 88A * 11.1V = 976.8W

I agree that doesn't sound like I/C fuel power density. 6 minutes of fuel driving 977W (about 1.3 horsepower) is probably quiet a lot less weight than 0.7 kg

Cliff

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
In other words say you combine cells in parallel (to increase voltage) and also in series (to increase current sourcing)

You want to try that again? ;) [last time I checked the universe worked *slightly* differently]

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Quote:

I don't know how much energy can be taken out of the LiPos.

As I say - cuurent design can drop up to about 20C.

I was talking about energy, not current. Is a 1800 mAh cell really a 1800 mAh cell or does is depend on things like temperature and current? I'd expect 1C discharge current to result in a different "usable capacity" than 20C.

I tend to post off-topic replies when I've noticed some interesting detail.
Feel free to stop me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

You want to try that again?

gitch,

Suggest you read rcgroups.com nSmP is pretty standard for LiPo use.

You use nS to get to the voltage you need (n=1 => 3.7V, n=2 => 7.4V, n=3 => 11.1V, n=4 => 14.8V)

Because LiPo aren't (or rather weren't) available in heavy duty amperage variants then to get more current sourcing capacity you have to parallel them up. So say you have got 3 cells as 3S giving 11.1V and they are 2200mAh cells being used at 10C so they can source 22A but you really need 44A then you put another 3 in parallel to the first 3 (6 cells in all) which are now 3S2P. That's tripling the base cell voltage and doubling it's current souring capability.

What about this is wrong?

As an example here's a 3S2P:

http://www.redrockethobbies.com/...

PS this page has some pretty pictures of nSmP packs (on the basis that a picture is worth a thousand words - maybe I just failed to explain what I mean?)

http://www.hooked-on-rc-airplane...

4S2P:

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The mSnP notation is quite common and it's also intuitive, but your description of that

Quote:
In other words say you combine cells in parallel (to increase voltage) and also in series (to increase current sourcing)

is not. That's what glitch was addressing.

I tend to post off-topic replies when I've noticed some interesting detail.
Feel free to stop me.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

wow, 6 rotors. Well that does it... I'm gonna invent something with 1 rotor and make everyone jealous. ;-)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Cliff,

I understand about placing cells in series and parallel very well... the statement I highlighted is backwards, you said connecting in parallel increased voltage, and series for current, when it fact it is connecting in series for voltage, and parallel for current.

Writing code is like having sex.... make one little mistake, and you're supporting it for life.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Totally awesome! This is my inspiration for my next projects!
Do you know who is the manufacturer of those Roxxy 2827-34 motors? (I'd like to see the datasheet)

I suppose motors are controlled without hall sensors..
(by looking at the pictures i see only 3 wires)

That would mean they are controlled by detecting the back EMF spike?
Do you know is it a sinusoidal, or space-vector control, or just six-step commutation?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Robbe makes the Roxxy motors, datasheet is here: http://data.robbe-online.net/robbe_pdf/P1101/P1101_1-477834.pdf
They are usually controlled by six-step commutations drive.
It's a sensorless design (no Hall-sensors) and the back EMF (BEMF) is used to control the timing.
The BEMF is measured on the inactive winding in each step, the timing of the zero-crossing noted and the next commutation timed accordingly.
Lots of info here:
http://www.gobrushless.com/kb/index.php?title=Brushless_101_-_Chapter_1 and here: http://scholar.lib.vt.edu/theses/available/etd-09152003-171904/unrestricted/T.pdf

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

nedward wrote:
I'd love to build one of these but looking at the site's shop page, it all cost WAY too much money!
Yes, that's why you should build it yourself, not hard and a LOT cheaper.

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Quote:

I expect about 15-20 minutes on a 1800 mAh 3-cell Lipo

20 minutes is 3C so you are planning to pull 5.4A from the cells. 3S is 3 * 3.7V = 11.1V. So that would be 59.94W. Usually that's enough to lift about 50..100 grammes in a hovering device, how much does the camera/mount weigh?

Camera is a Kx-181 HQ, weighs 30g. Tilt and pan mount weighs 20g including servos. And the transmitter weighs about 5g. Total about 55-60g.
I will be saving some weight on the airframe which is made of carbon-fiber instead of aluminum as the mikrokopter.

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Amazing sound, hard to beat.

But, you can go pretty fast on electrics too.
http://www.youtube.com/watch?v=q...

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

statement I highlighted is backwards, you said connecting in parallel increased voltage, and series for current, when it fact it is connecting in series for voltage, and parallel for current.

Oh sorry, I completely missed the point you were trying to make and my silly typo :oops:

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hehehe, we all make them. I chalked it up to a lack of coffee, or one too many pints ;)

Writing code is like having sex.... make one little mistake, and you're supporting it for life.