I am a complete newbie to this forum. I have not flown any RC aircraft yet but am very interested in avionics and robotics. I have flown a Cessna solo using VFR so I understand basics of aviation. I am software engineer so am handy with coding . I have some experience with soldering and electronics but it was a while back. Came across Chris's article in Wired and have got this major itch to make my own quadcopter :) I have a few questions for the community -
1. Is it better to get some experience in RC flying or would you recommend directly trying to fly a quad using Autopilot? My interest is mainly in autonomous flight at this point not the aviation per-se. What do you think.
2. Sparkfun has an autopilot board for http://www.sparkfun.com/products/8785 which costs about 30 dollars. http://code.google.com/p/arducopter/wiki/APM2board costs 200$. What are the differences between them and what would I need for a quadcopter ?
3. If the above question sounds stupid, please point me to some websites or articles that I can read to understand the difference.
4. I don't have a lot of cash to burn but can spend around 500$ on this project? Is this doable? If not , what is realistic considering my experience level?
5. I am upto any challenge related to learning, soldering , assembling , programming etc. Can I bring the cost down by doing some of these things myself?
Any help will be greatly appreciated.
P.S. I live in the San Jose , CA area in case that matters
Welcome aboard Amit, I would definatley start HERE. Everything you will need is in the Wiki. The DIY Drones store HERE has frames, electronics, etc. as well as complete kits, so check them out for pricing. You can also build your own frame to save some money but the kits have been tested and there is a lot of support on these forums relating to them.
Hope this helps.
I'm in the same boat you are with the "quadrotor itch". Most of my experience is with electronics and I'm working on some User Interfaces in Processing for controlling a quadrotor using a Logitech Dual Action joystick controller or a TouchOSC layout running on my iPod Touch.
Anyways, in regard to the two boards: the $200 ArduPilot Mega is completely furnished with sensors such as an accelerometer, gyro, magnetometer, pressure altitude sensor, and GPS. It has an atmega2560 which has 4times the power of the atmega328. The $25 sparkfun board is running that atmega328. It is similar to a barebones arduino pro, it has a port to plug a GPS into, place for servo headers, etc... Bottom line it requires the purchase of lots of additional hardware, soldering, and is not a ready-to-plug-into-quadcopter solution.
As you can see here: http://www.diydrones.com/notes/ArduPilot/ The $200 is a newer version of the $25.
I'll let the experts handle your remaining questions; I'll stick with my experience in electronics for this post.
Thanks for the info. That helps clarify many of my questions.
I started my Arducopter experience in a similar way. I had some basic RC, electronic, and soldering experience from building 120 pound Battlebots in school but no experience with any RC flying platforms such as a quadcopter. After browsing around the forums and Arducopter manuals on this site, I decided to build my own quad from scratch. After a few crashes and playing around with the PIDs, I'd like to say that the time and money investments in this project were completely worth it. I came across many skeptical RC hobbyists on various forums when I started my second RC project: an FPV Ritewing Zephyr platform. People told me that (with 0 experience in RC planes) I should start with a smaller glider, go to flight school, etc. And while I appreciated all of the helpful criticism, I was too eager (and maybe dumb), so I started flying the zephyr in a field near my home. One year later, I have only had 1 crash with virtually no damage and many successful hours in the sky. Moral of the story is that if you have the curiosity and interest in something, by all means GO for it! Just be SAFE too and be prepared to do a lot of repairing.
I agree with Ray K on his opinion with the two boards. You should go with the APM2 as the arducopter code is written and updated for the APM2. As Ray said, Sparkfun's Ardupilot is the original and less-powerful version of the modern APM2.
$500 may be a bit low considering you need to purchase an RC transmitter, batteries, battery charger, the APM2, frame, motors, ESCs, props etc. But it may be doable.
I'm just typing as I think now, but If you buy a
$50 Turnigy 9x or other cheap hobbyking transmitter/receiver (http://www.hobbyking.com/hobbyking/store/__309__190__Tx_Rx_Systems_...)
~$100 for 4x brushless motors and 4x turnigy 20amp ESCs (http://www.hobbyking.com/hobbyking/store/__2163__TURNIGY_Plush_25am...)
~$50 for a LiPo and charger
~$40 for a quad frame (http://www.hobbyking.com/hobbyking/store/uh_power_search.asp?idCate...)~$10 for props and spares (http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=...) It may be able to fit your budget. As for bringing down costs, I think the APM2 is the same cost with pins not soldered or pre-soldered. If you buy everything else above from hobbyking or other cheap outlets, this would save you money. I may have missed something though, it is very late!! or should I say early morning.... Good Luck!
I'd like to kill several birds with one stone here by discussing some of the ideas I had for my first quad in hopes of both providing the original poster with some ideas to consider and also getting feedback on them from other people: I do not mean to hijack this thread.
I can avoid buying a transmitter/ receiver altogether because of my aforementioned joystick --> laptop --> xbee plan. The APM 2 requires an xbee anyways if you want to have in-flight telemetry, and it has joystick support: http://code.google.com/p/arducopter/wiki/Joystick
Making this quad requires purchases from several different places. Hobbyking is good for the frame, motors, ESCs,
The store here has got some stuff. I was thinking of making the Hobbyking order first. I already have plenty of electronics equipment around the house/ workshop (the two become synonymous sometimes) including an Arduino Pro Mini - so once I build the mechanical portion of the quad I can start doing some basics with that and perhaps an acccelerometer/gyro sensor board from sparkfun. Can anyone comment on whether this one: http://www.sparkfun.com/products/11028 will work sufficiently for a quad? I know it doesn't have a magnetometer, but if I'm not using GPS I don't need that, do I?
That's where I plan to start. Afterwards, once I can see that the frame is working, then I can do one of two things, I can keep adding sensors manually, or I can get an APM2. I can tell you that the $200 price is cheaper than buying sparkfun modules individually, but more expensive than if you have PCB experience and can put the same chips on your own board. For example: the APM2 consists of:
So, the $200 all-in-one, has more features, hassle free, guaranteed to work out of the box board is actually quite appealing now.
Back to Hobbyking, this is what I'm planning for my quad:
Frame: X525 Frame Kit
I'm having a little trouble finding motors and props. In the frame product description, they say
"28-XX 1100~1300KV Brushless motor x 4
8x4~9x5 Propeller - 2 standard/2 reverse rotation"
Does the motor - 28 mean that the inrunner is 28mm?
And I could use some help deciphering the prop specs.
Still working on figuring battery stuff out too...
Again, didn't mean to hijack this thread. My concluding remarks are that I'm thinking about a step by step build to not have a sudden surge of spending. Hobbyking order first: frame, motors, ESCs, props, and a battery. Test frame with Arduino I already have I may place a sparkfun order for IMU board and XBees since you need to get the XBee radios separately anyways. Then, I might purchase the APM2 here.
Amit, I hope this helps. Everyone else, please comment and help the both of us.
@ Ray K,
I would not recommend going for the [Joystick > Laptop > Xbee] setup for a primary control system for a quad, at least not for a beginner and a quad that may NOT be fine tuned just yet. Reason being is that the latency is a bit high compared to a regular RC Transmitter. But, in the spirit of my previous post, if you want to try it in place of a regular Tx, by all means, go for it. I'd love to see the processing/touchOSC system integrated as well.
As for the telemetry, you can also purchase the 3DR radios (in lieu of xbees) which were designed to integrate with APM and mission planner. Here's a link if you're interested: https://store.diydrones.com/3DR_RadioTelemetry_Kit_915_Mhz_p/kt-tel...
I just don't quite understand your methodology regarding building your own control platform and then buying an APM2. Why not just save the money and just buy the APM2? Building your own microcontroller will require hours of research and coding and you won't know whether problems exist in your controller or in your frame + motors since there are 2 untested variables. Like you said, the APM is all-in-one, much less hassle and guaranteed to work. Just my two cents.
As for the battery and props, I guess it depends on your frame and motors. I'd probably go for a 3s LiPo somewhere between 2-4 amp-hours capacity??? I would go for the props that they recommend and try those firstt :)
Hope this helps :P
Helps quite a bit, thanks.
I see why you don't understand my methodology regarding the APM2. All I was getting at is that I want to play around with the frame I build using an arduino I already have before making a $200 investment in an APM. As for purchasing a gyro/ accelerometer to facilitate this, your point is correct.
This is the first I'm hearing of XBee latency, but after comparing specs between the 3DR and various Xbees I see what you're talking about.
I was just letting you know about the 3DR radios in case you haven't heard already. The xbees work perfectly fine too! Its a matter of personal preference.
In my experience with joystick control, there is a large fraction of a second delay between my joystick input and the quad's respective response. This can be quite scary when your quad is plummeting to the ground or heading for a tree!!!! But it does work and is a really cool feature that you should try nonetheless. I was also trying to give Amit some extra information on the idea you brought up so he can make an informed decision.
As for the source of the latency, I'm not sure if there is a noticeable difference between xbee and the 3DR radios, but the whole process (Joystick > Computer > Mission Planner > Telemetry, etc.) does create this delay I am referring to.
I understand. And, now that you mention it, I imagine I'm going to have even more of a latency issue between TouchOSC GUI running on iPod Touch --> Router --> Processing Sketch running on laptop -->Xbee/ 3DR radio telemetry --> quad, which was the original idea.
Yea I agree. Hmmm, now that I think about it though, the AR Drone has a similar control interface to what you are describing above involving an iDevice and has RC-like responses. But then again, the ARDrone's range is also limited by WiFi
The AR.Drone did come to mind as I was writing that last post. Wifi is much more capable than any of the other telemetry options we've discussed. Plus, AR.Drone has cameras and costs only $300.
This is the ultimate issue. Xbee gives us a range of miles, but implementing FPV makes life a lot harder. AR Drone, on the other hand, is limited to the range of wifi as you said, but has phenomenal video streaming capabilities.
What we ultimately need is to mount a smartphone on the bottom of the drone with the back camera facing towards ground. That hardware takes care of telemetry, video streaming FPV, IMU, GPS + Compass, and a powerful processor for higher level "thinking". A lot of people say Android is well suited for this, but I'd shoot for a Jailbroken iPhone. My phone is not a smartphone, and I've always said that if I'm getting a smartphone it would be an iPhone, and I jailbreak, Also I'd need to get a Mac to do application development, even for Cydia or local deployments. My current attempts to circumvent the Mac requirement for iOS App Dev are not going well. But that's a different story entirely. So, we really need to mount a smartphone on a quadcopter and write an App that combines all of these features. A backup autopilot would be key though. I'd still want a hardware IMU and GPS onboard in case the phone died or dropped, and there's also the possibility that iPhone serial can't provide high enough resolution for spontaneous movement of the quad.
Thanks much for the advice.