Project log: Ducted fans hovering craft for air photography

A

avner

Senior Member
Hi,
This will be my project log for this imaginative device.

Goals:
to build a flying device
to be able to control it from ground.
to be able to mount a DSLR camera on it for air photography.
to be able to control the camera - including tilt and swivel.
and the most important - to be able to land it back safely

I will use:
Multi-picaxe based system to control the servo's and motors ESC
Accelerometers
ultrasonic distance meter for gentle automatic landings
maybe air pressure sensor for altitude metering (will replace the ultrasonic for landing)
A 2.4 'spy camera' will be mounted on the DSLR visor, sending the image to ground, where i'll have a monitor to see exactly what i'm shooting.

Driving power devices:
2 or 3 100mm duct fans (2.8kg thrust each)
100A ESC's
additional motors to give horizontal thrust and direction.

Communications:
Hopefully having enough speed and reliability with HM-TR HopeRF wireless modules,
otherwise I might decide on going wired for both power and comms.
Yes, I might end with having some 30m of 2-wire power cable + 2 optical fibers for laser comm.
the craft being wired is not too bad for me, it will allow me to connect to a much larger battery for very long flight times, and i need it mainly as a vertical elevator for the camera, so wired is acceptable.

the gear is on its way, i will post as soon as things arrive and being experienced...
 
M

manuka

Senior Member
What is your budget ? You've not mentioned ultimate intended range or altitude, but if that 30m means "nearby" I'd say initial experience with Li-Po FPV (First Person View) model planes at ~2.4 GHz is now worthwhile. 433 MHz probably wouldn't have the bandwidth or power for your needs. Check the pix below of a local here in NZ- his fly by aloft video is astounding, yet the setup was "very reasonably priced" and ultra reliable.However he'd put in many hours learning to control the little darling.

Small helicopters (especially quadrotors) are traditional for fixed images- real estate agents find them great for interesting views when marketing property etc. Is that your intention? I'm surprised the Japanese didn't consider using them for inspection of those pesky reactors in fact !

Now why would you send a heavy DLSR aloft? All manner of dirt cheap light weight 10MPx point & shoot digitals exist now. You may get better shots just using one with image steadying features. Stan
 

Attachments

Last edited:
A

avner

Senior Member
Sure i can use smaller craft with smaller camera,
but I plan on doing some videos as well, I use the 5D mk II.

also, not being a super-pro RC jet pilot,
my intention is to make a device which is safe, sure, slow-moving and easily controllable.

why 433 MHz wont be enough ?

I plan on transmitting quite a small amount of data.
the craft will be half automated, meaning it should hover in place by default,
the TX-RX commands will be "move down\up" "move left" etc....

that what i mean when i say a 'slow moving' craft - it doesnt need comms for realtime flight commands.
 
S

S2L

New Member
Before you start take a look at Quadrocopters or hexacopters. I would be tempted to go down the traditional route of using a receiver, speed controllers and a gyro then control everything else with picaxe, unless of course you are doing the whole project as a picaxe challenge

Edit, if you want stable you want to go big, say 400 to 500mm span minimum
 
Last edited:
R

rgooge

Member
Try one of these...

http://ardrone.parrot.com/parrot-ar-drone/usa/

with one of these attached...

http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=150564145350&ssPageName=STRK:MEWNX:IT

using the cam like these guys do...

http://www.rcgroups.com/forums/showthread.php?t=1362692

.. this is what I've done. The stills from this little keychain beast are good enough to read a number plate on passing vehicles when used pointed out the windscreen. You just still frame what you want. I use the FPV on board for alignment and have the HD video running the full length of the flight.

OK, you want hi-res stills, but this is relatively cheap platform (maybe the need to purchase an iPod Touch to control it). Oh, and you'll have to forgo the joys of building it yourself!!

What you're trying to achieve is pretty difficult, controlling the drone as well as lining up to take a decent photo.

It's a decent project to embark upon and good luck with it.
 
M

manuka

Senior Member
Avner: I'd suggest you pop along to a nearby model plane outing & get insights on typical gear now in use. This is only a field I look at occasionally, but I've been gobsmacked with the sensational developments that have occured in the last decade! Aside from it's tight bandwidth 433MHz is legally allowed only 25mW in most countries- reliable ranges are likely to be hence quite small, & nearby signals may override yours.
 
D

Dippy

Moderator
Let me know how you get on.

I want to do a similar device with video but I'm it'll look like a UFO.

I'm just trying to design a BLDC driver board and software at the moment. I'm planning on Maxon motors.

For the radio part I'll be making a bespoke design in the 868MHz region.

I'll be interested in how you grapple with the mechanical section really.
I'm leaning towards vectored thrust and make the device as autonomous as possible.
 
D

Dippy

Moderator
No, but if you see the headlines "Idiot from Dorset, UK arrested for 2metre UFO" then that'll be me.

I'm looking forward to the build but it'll be long winded and expensive (predictably) as there will be nothing nasty in the construction apart from my soldering.:)
It'll probably never get off the drawing board , let alone getting off the ground.
 
M

MartinM57

Moderator
I'd recommend a lot of test flights with a house brick as a payload rather than your 5D MkII - they probably don't like being dropped that much ;)
 
slimplynth

slimplynth

Senior Member
Dippy said:
Let me know how you get on.

I want to do a similar device with video but I'm it'll look like a UFO.

I'm just trying to design a BLDC driver board and software at the moment. I'm planning on Maxon motors.

For the radio part I'll be making a bespoke design in the 868MHz region.

I'll be interested in how you grapple with the mechanical section really.
I'm leaning towards vectored thrust and make the device as autonomous as possible.
Click to expand...
Hi Dippy.. just googling my life away but have you seen the below elsewhere on the net? may prove useful..

http://jlnlabs.online.fr/gfsuav/gfsuav.htm

reminds me of this story..

http://www.dailyrecord.co.uk/news/weird-news/2011/03/03/revealed-the-day-britain-was-put-on-alien-invasion-alert-thanks-to-student-pranksters-86908-22963298/

except yours might actually fly.. all you need to complete is to have a shaved monkey as the pilot.. just don't fly near Hartlepool ;)
 
D

Dippy

Moderator
That's fancy.

No, mine will be working on the Comeonyoubastward lifting principle.
Oh, it'll end in tears I'm sure.
 
A

avner

Senior Member
hey guys, back to topic....

I've played a bit with the HM-TRs
each with an 08M @ 8 MHZ, 4800bps

seems like it has a rate that will allow me to send the wanted data about 10 times a second...
i'm talking generally, only estimations.
i guess that 10 bytes would cover all my communication needs...


Click for the test video

.
 

Attachments

Last edited:
D

Dippy

Moderator
OK, so you can make it twitter at 2 cm...
So far so good.

But what range do you want?
What happens if it goes out of range or is interfered with?
 
A

avner

Senior Member
I've checked with the longest path and behind many walls in the apartment, but it never lost signal.

I have to check it LOS, because the craft will always be on sight.
tomorrow i'll take it for a little trip around the block
 
M

moxhamj

New Member
This sounds a fascinating project.

I made a picaxe controlled car using RF modules once and it does work. The catch is that running servos and serin on an 08M resulted in occasional glitches. So it might be better with a picaxe dedicated to radio, then maybe use another method of getting the data to the servo/ESC picaxe (eg an analog voltage, rather than serin). The newer generation picaxes may not have these problems.

There are some excellent comments above. I agree with the comment about cameras. I put a keychain camera on a model chinook once http://www.youtube.com/user/DoctorVernAcula#p/a/u/2/igs_P5cyvls

I learnt a lot from that. A chinook has a lot less vibration than a helicopter with a single rotor, and a quad or hex copter would have even less vibration.

My major issue is I live on top of a windy hill. Flying things that are ultra stable (counter rotating helicopters) don't have enough forward speed. Flying things that can counter the wind are more unstable.

It helps to learn to fly on cheaper things like indoor helicopters. 4ch ones are a lot better than 3ch if you want to learn how to fly a quadcopter. http://cgi.ebay.com/4ch-rc-Z008-Avatar-2x-L-R-side-motor-Heilcopter-PARTS-/390292320140?pt=LH_DefaultDomain_0&hash=item5adf3c2f8c

Some quadcopters make it look easy http://www.youtube.com/watch?v=TLKWjUN9JHs but I know from experience that this pilot is *very* skilled!

Re the time constants for using an RF module, I'd say that it probably is ok that it is a little slower than a standard RC link, because if you are needing to update every 100ms you are probably going to crash if you are a beginner! The indoor ones manage to hover precisely with no changes to the controls needed, so it ought to be possible outdoors with maybe GPS and small changes to controls to counter wind drift.

With a heading gyro, ciompass and GPS, it may well be possible to control this with picaxe.

Keep us posted, this sounds like an interesting project!
 
W

WHITEKNUCKLES

New Member
Avner
From your original post

Sorry to put a damper on things, BUT

UAV
First investigate local Unmanned Aerial Vehicle Regulations.
These can be quite Draconian and have the force of law but are not easily found.
Civilian Autonomous Flight is not normally allowed.

Manchester Police were grounded by the CAA for not having a registered vehicle or a certificated operator.

Then find insurance.

For your needs you could Google ~~
Kite Photography
Balloon Photography
Mast Photography, your best bet.

Pole photography is on a smaller scale but is surprisingly successful giving an entirely different perspective for a small, say less than 24 foot elevation from a Carp Fishing pole.
Cost can be very reasonable if one can improvise.
Panning can be done by rotating the pole.
Picaxe can easily handle Tilting and 2 stage shutter control by wire or radio control.
Useful for archaeology, monuments, fairs, exhibitions and of course house sales.

The novelty of the improved perspective makes such photographs eminently saleable. :D :D

Dave
 
A

avner

Senior Member
C'mon... you dont want that project to stop.. do you ?

mm,
anyways, i just ordered joysticks for the controllers.
great price for that cool analog stuff ! :D
 

Attachments

slimplynth

slimplynth

Senior Member
Hopefully agents from Big Brother aren't reading this thread and tracing you as we speak :)

I was thinking about insurance issues... assuming you're a sensible person you'll find somewhere quiet to do testing...

If it were my pennies at stake I'd want to experiment with communications that were excellent/fast/reliable at short range.. to guard against damage to the craft perhaps have beacons set at a decent range, to act as a kill switch fence.. perhaps bluetooth to control the craft and 433 for the beacons... the beacons might send a very simple repeating signal that a picaxe would catch and follow a safe land subroutine on receiving?

Just an idea.. might also save some frantic running as you watch your pride an joy get swept away by the wind... it all happens very quickly

I got a cheapo 3Ch RC helicopter for Christmas.. Christmas day I just watched it go above house height.. it's coaxial.. didn't know that trying to make it go forward would cause it to go higher... 100m later.. and a small dot in the sky it auto-ditched; i was fortunate enough to be able to ask for it back out of this guy's garden.. felt like being a 12yr old again.. better that than explaining to someone why it had scratched a beloved car.. lesson learned...
 
A

avner

Senior Member
LOL,

That's another reason why i keep an eye open for a wired version !!
at least i have a limited cable length...

second thought, i'm not sure i'll be able to go much higher than 30m with cables.
imagine the weight of 30m heavy power cables + optical fibers.....

anyways, that's what i mean when i talk about stable and slow.
even if you wish - control will limit you from doing sharp and fast moves.

and about the wind... i'm not going to go GPS. thats too much.
but i have a great solution - wind-free photoshoot days ! ;)
 
slimplynth

slimplynth

Senior Member
as digital memory is a plenty.. how about a photo-cannon :) a parachute... ooh that makes me think of those aerofoil kites.. the ones that inflate like a wing.. have mega amounts of lift for a small form.. possible candidate for a stable picaxe UAV? not sure how good a picture you'd get from a constantly moving platform.. but some lighting effects might give the illusion of a UFO Dippy?
 
N

NXTreme

Senior Member
slimplynth said:
ooh that makes me think of those aerofoil kites.. the ones that inflate like a wing.. have mega amounts of lift for a small form.. possible candidate for a stable picaxe UAV?
Click to expand...
Something like this, right? Getting it to fly just right on autopilot might take a bit of work, but it'd look beautiful once you got it going!
 
M

manuka

Senior Member
Avner: Before this spirals further, can you please clarify just what you are most intending to photoshoot! Houses? Birds nests? Crop circles? Nuclear reactor cores? Security?

Although I'm guilty of promoting these HopeRF transceiver modules some 3 years back,times have moved on & in any case I'd have significant reservations about using them in demanding applications. They're certainly a cut above el cheapo garage door opener offerings (& ideal for terrestrial hobbyist type needs),but I've visions of all manner of mischief arising if they misbehave aloft. As others have mentioned, you could be setting yourself up for a "punch on the nose"...

Footnote: Ensure you tick all the boxes at =>http://www.rcapa.net/guidelines.aspx
 
Last edited:
A

avner

Senior Member
somehow i'm not too afraid of the HM-TRs abilities.
i could get a few hundred meters range with it outdoor with obstacles.
in-flight there should be no obstacles, pure LOS, and with antennas well above ground
all that said without being a RF pro....
i would possibly like a better range and safer transmitters.

i'm willing to spend 100-150$ for a pair of TX and RX,
but I need it super simple, transparent link, without programming, and serial, like the HM-TR
can you recommend ?

the use will range between photography of houses for real estate, parks, bridges and civil engineering projects, etc...and of course the joy of building....
 
A

avner

Senior Member
I've just looked into the Xbee pro (60mw)
it has a range of 1500m. not extremely longer than the HM-TR.. slightly double - which i reckon, is worth the effort.

it also has plenty of picaxe tutorials and stuff so i probably won't go messy with SPI I2C etc....
might be the way to go....

yet, i promise myself that i'll get it only when the craft first takes off with the HM-TR. :D
 
A

avner

Senior Member
OK guys -
plans changed.

I'm going to make this a QUAD project.
4 rotors, instead of ducted fans.

the reason is that with ducted fans, all spinning the same direction - i will have forces that will cause the craft to spin as well. (like a helicopter without the tail prop.)

in the quad design, it will have a '+' shape, with a rotor at each end,
i'll have 2 rotors turning clockwise, and 2 rotors turning counter clock wise.
that way it cancel out the spin forces and stays aligned.

i have already ordered special props, 2 CW props and 2 CCW props (one pushes air, and one pulls). of course that the CCW props will be mounted on motors that spins counter clockwise too.
that way after all - all motors will be pushing air downwards.

well, hopefully :D

i thought of using that big @ss ducted fan as a boost engine in the center of the quad, to give it extra lift force for vertical ascending....
 
M

moxhamj

New Member
Quad sounds a good idea. You will get less vibration and that is important for video.

Re RF and going out of range, well that can happen with any radio controlled toy. I wonder if you could program it to go to a 'safe' mode if no data arrived, or lots of packets were coming through with corrupted checksums.

'Safe mode' could be as simple as a hover at the same altitude, or a very slow descent. It could be useful too for when the batteries are about to go flat.
 
A

avner

Senior Member
Thanks for the link,
what a pity, they have 14" props, but i've already ordered 12" ones.....
anyways in case i'lll need it :)
the rest is not too interesting....

DrAcula,
it's exactly what i've been thinking.
but ive gotta get a barometer for that...
anyways i still have time..

here's the flowchart:
 

Attachments

A

avner

Senior Member
Nice link. gives some hope !!

i know of the ArduCopter, i'm ordering part of the hardware (props and motors) from them.
but my challenge will be the picaxe programming, as i'm not a programming maniac !! :(
 
A

avner

Senior Member
MartinM57 said:
Have you verified that a PICAXE (or a number of them) can be coded to control the flight of a quadcopter?
Click to expand...
why not ?
hey, i'm doing it from scratch. i'm independent of other modules.
yes it will be a challenge, a complicated one.

in the control phase all you need is to control the rotors.
which means control the ESC's - they use the standard picaxe SERVO command....
 
M

MartinM57

Moderator
avner said:
why not ?
Click to expand...
Maybe (and I don't know) the limited mathematical capabilities and relatively slow execution speed will be showstoppers in implementing the responsive flight control algorithms required for fundamentally unstable airborne craft?
 
A

Andrew Cowan

Senior Member
I know that alot of the people who make quadrotors have to make major adaptations to the brushless speed controllers. The reason for this is that normal r/c speed controllers take a PPM input, which updates every 20mS. This is really too slow for a quadrotor, which needs to respond much faster, otherwise you'll get lots of overcorrection and oscillation.

Most quadrotor speed controllers have been adapted to I2C control - at 400kHz, this is much faster, and a quick enough response.

I very much doubt that a PICAXE would be fast enough to run your quadrotor - although the PICAXE would be fine for giving general commands to a slabilisation system (such as a few gyros).

In addition, commercial brushless speed controllers have limitted ramp up/down rates - you need very fast responses and dynamic braking. Without serious modifications, the ESCs you have bought are likley to be too slow.

http://www.rcgroups.com/forums/showthread.php?t=958957 is the sort of response rate you really need if you want to use the quadrotor for anything more than indoor hovering.

If your control loop isn't fast enough, or your speed controllers are slow to react, your platform will be unstable, like http://www.youtube.com/watch?v=490BUkMnPk4&feature=player_embedded.

Andrew
 
Last edited:
You must log in or register to reply here.
Top