/diy/ - Do It Yourself

Picture of the board.

>>8038
How much has that cost you so far, OP?

>>8098
Hmm... I ordered a bunch of other components and tools along with it, so I don't have a definite number, but maybe around $80, or a little less?

Props: ~$4
Motors: $25
2 LiPo batteries: $8
Carbon Fiber: $5
Controllers: $25
Board: Free (~$10 for 10 on Seeedstudio if you want them made nicely)
Components: $13 (Sensor ICs sampled off STMicro)

I've purchased a ton of other tools and unnecessary things along with it which probably boosts the total cost to around $160-ish. Stuff like a LiPo charger and some overpriced sensor breakouts that I didn't need from SparkFun.

>>8172
That's awesome

Have you thought about adding a camera to it?

I wish you would resize your images, especially since 4chan has the bandwidth of sticky molasses.

>>8184
I bought a cheap, small, camera that I'm thinking of attaching to the bottom, but I'm currently aiming higher with thoughts of a making a cheap scanning LIDAR to attach to the bottom to do 3D mapping and autonomous navigation.

If you're subscribed to IEEE, look up "A Low-Cost Laser Distance Sensor." Same idea is applied here: http://www.parallax.com/StoreSearchResults/tabid/768/txtSearch/laser/List/0/SortField/4/ProductID/77
4/Default.aspx , except in the paper they spin that shit at 30 Hz with single degree resolution.

The biggest problem will be attaching something that spins on a small UAV like this... Can't find any cheap sliprings that are small enough, and then there's the motor and the oscillation it'll create. I might just end up putting IR distance sensors on a micro servo.

>>8224
Why does that laser range finder have to spin?

Completely new to this, by the way.

Also, how do you control this?

Is there antenna, controllers, etc that you need to attach?

Use Ardupilot!

www.diydrones.com

>>8259
I mean, you could fix the range finder and not rotate it, but the point is to get high resolution depth data at high rates.

Current revision of the board has no wireless communication, but next version will probably have a TI CC1101 on it. That way, I can get communication with http://processors.wiki.ti.com/index.php/EZ430-Chronos?DCMP=Chronos&HQS=Other+OT+chronoswiki , not to mention, I already have a CC1101 dongle for my PC.

>>8298
So how do you control it now?

Just connected by a wire?

>>8321
Well, it doesn't fly right now, as I designed the board wrong and the microcontroller isn't responding. I'm redesigning it and having it sent off to be made.

>>8332
Be sure to make a thread once you get the new parts.

>>8332
Ah that's a shame.

>>8350
This

>>8371
Will do. The only real problem with it was that I left one of the reset pins floating, but there are some other minor errors to fix.

>>8395
>I left one of the reset pins floating

What's the problem with that?

>>8407
Well, it's not actually the reset pin, but the brownout enable pin. Datasheet says it wants to be held either high or low, and leaving it floating probably leads to some sort of startup issue.

>>8395
do you use eaglecad?

care toshare any tips for design?

>>8447
Oh, OK.

Well I wish you the best of luck man, and look forward to your future threads

>>8464
I do use Eagle, but I can't say I'm particularly experienced with board design, so I can't offer many tips... Just keep clicking that ratsnest button. Ground planes are cool, only resort to thinner traces if you absolutely need to, and don't solder on a QFN IC upside down like me.

>>8487
how do you prep the file to send to a board maker?

>>8500
There's a good tutorial on SparkFun explaining the process and settings they use to generate gerbers on Eagle that they send off to BatchPCB.

>>8297
What? And ruin all the fun bits? Plus, pseudo-C = shit.

Oh god, I'm so glad people like this are on diy.

I've done a few stupid projects with my aurdino, how much harder are drones then basic autonomous rovers?

>>8535
Still, ArduPilot is quite well made. I think they implement a Kalman filter and all that jazz.

>pseudo-C = shit
Oh boy, wait until you see Spin programming language on the Propeller.

>>8546
for most people, the hardest part is either debuggin the construction or doing the logic.

people tend to be hardware or software oriented.

>>8546
The biggest difference is that flying robots need to stabilize themselves and not just crash and burn. Quadrotors are probably the easiest to start with.

Google search mav-blog kalman filter of IMU data - it explains it beautifully. Unfortunately, it looks like his site is down at the moment.

A simpler filter to use than a Kalman might be a complementary filter, which I hear runs quite speedily on Arduino. Take a look at the ArduPilot code and see if you can make any sense of it.

Thanks for all the help anons

>>8552
Looked at an example and OH DEAR LORD!

bump

Simple question. What do you have to know to even attempt at creating something like this ? And how long did it take you to assemble it ?

>>8826
>the skill to fly/steer an RC vehicle
>understanding of electronics
>soldering skills
>free time
>patience
>money

>>8826
I learned most of this stuff as I went. A basic understanding of circuits helps. If you want to go the AeroQuad route, most of the coding and electronics has all been set out for you - all you need to do is build the frame.

Electronics probably took longer for me than the hardware, and I'm a software guy. Took me maybe 3 days of work to build the frame at a local shop. Obviously is higher if you want to build something larger than what I did.

>>8866
I always thought the idea of a flying minion was cool. If only I had the time to write the code for something like that. =/

>>8840
>>8866
I'm just starting to get into programming in my university. This type of crap is years beyond me.

>>8949
>yfw I'm in high school
The internet is your friend. Also lots of people on IRC willing to help you out, most likely

>>8949
If you have the drive to learn, the only other real obstacle is money. Since you're going to university I assume you have that.

>>8038
Awesome, this is what DIY should be about. So does it fly yet? Have you got your control algorithms worked out?

And why don't you have GPS, proximity sensors, or a way to manually control it?

>>8552

Yes, an EKF is really essential for AHRS stuff. Sparkfun has a nice module that implements this, but it could be fun to try it yourself. Start with a Baynesian filter and go from there.

Also, if you plan to fly it outside, definitely think about GPS!

>>8981
Read the thread~
>Doesn't fly yet- control board is busted
>Control algorithm- Kalman filter running off gyro/accel
>GPS- Probably too small to be outdoor
>Proximity Sensors- IR distance or scanning LIDAR
>Manual Control- will be implemented once I have wireless communication. I have a TI Chronos and a PS2 controller breakout that I'm itching to use.

op, how did you do the soldering? I've never done smt

>>8977
Nope I'm an extreme poorfag. I was just lucky enough to get government funding. My biggest obstacle will probably be electronics and hardware.
>>8970
Awesome. Where I live it's considered a Masters final exam to build something like this.

>>9040
My first time doing SMT as well. I used the frying pan method. It's actually not that difficult. Used a mixture of solder paste and regular solder+flux for some more delicate traces.

Man, this board is great.

>>9034
>Manual Control- will be implemented once I have wireless communication. I have a TI Chronos and a PS2 controller breakout that I'm itching to use.
Why not get a way to use a 2.4ghz hobby controller?
You can get some that have a 2 mile line of sight signal.

>>9135
Original plan was indeed to use Nordic 2.4GHz chips, but keep in mind that lower frequencies are the ones with higher range. Higher frequencies have less range with more bandwidth. CC1101 is sub-1GHz. The main reason is really just because I suck TI's cock and it might be cool to control a UAV with a watch.

I've always had a dream of building a long-distance UAV. Someday, someday...

>>8949
Dude, C is one of the easiest low-level languages to learn ever. Even moreso than Java and C++, and this is coming from a C++ fan.

>>8970
>>8949
Meet me on #robotics irc.freenode.net later on tonight. i will be __mike__

>>9182
Really? I had to learn C to do some CUDA stuff, and my god, it's basically just abstracted assembly.
I think it's an excellent language to learn and use for performance applications, but by no means is it simple.

Okay, the battery looks pretty heavy to me. Would this Quadrotor be able to lift it up?

>>9223
I've tested running the a *single* motor tied to the frame and battery, and it was able to generate enough thrust to lift them. Brushless motors really are something.

Hi OP,

Nice work you've got there. I'm jealous, been thinking of doing this myself.

Any /diy/er know where I can find brushless motors at a reasonable price? Each motor is like £20 wherever I look!

How OP got them all for like $25?!

>>8840
skill to fly a quad?
try a collective pitch heli
but a quad is way more sutible for FPV and such

^my cheap night heli

>>9350
Search hobbyking.

>>9352
Meh, I was thinking any kind of flying machine in general.

>>8840
>the skill to fly/steer an RC vehicle
tether it when you start practicing, and start practicing at 10% of nominal power. work your way up.

>understanding of electronics
Not much to it, really. I come from a software background and the logic of each piece is pretty straightforward. electromagnetic interference is where it gets funky for me.

>soldering skills
practice! the worst you can do is burn down your house.

>free time
Turn off your screen(s) and every other item in this list is more accessible.

>patience
When I'm working I'm too zen to remember things like food. Glad breathing is still in.

>money
Cancel your TV and stop drinking. Also, you might not want to wake and bake if you have to handle sensitive electronics. Reward yourself after it works.

But seriously... if you have all these excuses to get nothing done, WTF are you doing in diy?

>>9446
What's your point? That you can make things without time, money, knowledge, or experience/practice?

What motors is OP using?

>>9352
are there more rc pilots in here?
we might me able to help OP

@ OP looks a bit small to me
whats the RTF weight?
1S lipo?
im afraid it wont handle wind verry well...
just my 2 cents
it looks RTF
keep it up it'll be awsome

>>9568
I'm not too much of an RC guy- what does RTF mean? Real time flight?

I'm using 2S LiPos.

>>9587
RTF means, ready to fly

youre not an rc guy and youre building an quad...
very impressive!
btw: you might wanna learn to fly first
try a rc sim (phoenix RC has a gaui 330X quadro-copter)

>>9656
Bump from the depths of /diy/...
The idea is to use a filter and make it auto-stabilize, so that I just need to tell it what direction to move in. Is that plausible?