I'm a first year student currently studying Robotics with AI and for the past 3 or 4 weeks, we've been working with arduino clone boards (Seeeduino) with a custom shield full of inputs and outputs as designed and built by my university.
This is all well and good but I'm more interested in making things fly autonomously - which is the field I hope to enter in an R&D capacity when I graduate.
Not to snub some of the fun things you can do with the arduino but turning leds, buzzers etc. on and off and reading temp sensors is becoming a bit tedious. I do have some electrical knowledge behind me and a good bit of programming knowledge under my belt and I feel I want to take on something more challenging and interesting.
I find I can get to grips with and enjoy subjects far more if I can relate them to something I want to do and am interested in. Whether that be at the deep end or not, I find this best suits my learning style and it also motivates me to carry on when I'm stuck with something; unlike tedious exercises do.
So here I am looking to build a quad platform I can experiment with.
The problem is, I don't have a massive budget; maybe £200 or possibly £300 if the Mrs doesn't twig on LOL
Most of the ready to go kits I've looked at are seriously pricey so I figure that only leaves me with the option of building my own.
I have one or two ideas swimming around in the old noggin but I'm not sure which will be the best route to go so I thought I would pitch them against some more experienced Drone enthusiasts / experts.
I've got my eye on various parts to build my platform but I'm having a hard time getting my head around matching and pairing motors props etc and I'm also trying to work out the best option regarding a flight board.
I had my eye on the ArduPilot Mega v2 but that seems to be half my budget gone right there so I got to thinking about creating my own board based on the APM.
Not sure how economical that would be or how I would go about it in any detail yet though.
Going back to matching the parts up for my design, I've been using these two calculators to help me get an idea of whether or not the thing will even take off.
I've run a few motor / battery / prop combinations through the multi coptor calculator but the results back are hit and miss. You can't always find all the information about a particular motor that the calculator wants so it becomes major guess work.
Here’s my proposed design so far and without a proposed power source or flight control board.
Perhaps a more experienced person could give my specs the once over to see if there are any obvious floors or concerns?
Initially, I'm just interested in getting a platform in the air that is capable of taking off, landing, hovering and maybe following a couple of simple way points.
Total Weight So Far: 540g
Expected Final Weight: 1.1 - 1.3Kg
Total Cost So Far: £71.32
The one thing I don't want, is for the thing to barely be able to haul its own weight and has to hover at some rediculous throttle setting like 70% while sucking the life out of my battery at a rate equivalent to a 5 year old drinking juice on a hot day.
One other thing I'm not sure of is the need for spending out on fancy RC Controllers.
Since I believe I will only have need of manual control for the purposes of taking off and landing and while the vehicle is in close proximity, I'm wondering if I could just get away with pairing up any old cheap RC transmitter to an arduino capable receiver as long as the frequencies match?
Hope someone can lend a bit of advice for a novice and thanks for reading!
It is always a good plan to invest your time in doing something physical, rather than a purely intellectual exercise. The real world gives back real results and those are almost always quite different and unexpected as compared to just the calculated.
I can't give you all the answers you seek and what you ask opens more questions.
The HK frame you picked is on back order so I hope you are not in a hurry, plus the shipping cost. It is also rather small. That raises the question about your payload both weight and size. Beyond a controller what else do you plan to carry over the first year? You seem more focused in buying a frame and making a controller. Perhaps your priority is reversed? It is possible to build a reasonable size, strong and low weight frame for nothing or very close to nothing using simple tools and materials. If I was in your budget situation, I would spend some time looking over that aspect first.
In regard to the controller. I don't know what facilities are available to you, but making an APM is complex and certainly not a trivial exercise for anyone that is inexperienced in hardware creation. The APM2 for instance is small because it is a high density 4 layer pcb and well thought out design. Going to a single layer, double sided pcb, that most home equipment can handle, would make it very large physically and then there is a weight penalty. Going to multiple stacked boards is also problematic just to name one solution. Even if you go back to earlier versions it is still problematic to build yourself. Then there is buying "one off" parts, both in availability and cost. Often one off parts are simply not available and not all factories/wholesalers will give you a sample part. So depending on what free resources you have available to you, to build up the pcb/assembly, the cost would easily exceed the cost of buying a pre-built and tested APM2 and you would still end up with something far less capable than the APM1 or 2. Then there is the time factor. Do you want to spend your time on making a controller first or writing software? What is your time-line for this project? Graduation?
Here is what I got from you.
"This is all well and good but I'm more interested in making things fly autonomously - which is the field I hope to enter in an R&D capacity when I graduate."
This points me to thinking you want to spend the majority of your time on development of flight control, not hardware re-design.
The APM2 has just about everything you need and perhaps more already included and the price is excellent over anything out there that's even similar in capability. Plus everything is open source of course and not half and half like many other designs Then of course the firmware and past experiences are very well documented so that you could easily get into what you want to advance instead of spending months re-writing the wheel.
There is a very important factor here. It doesn't matter what your budget constraint is. There is a point where it just isn't enough no matter how you scale down and scrounge. My opinion is that your budget is on the teetering edge for what you hope to do. So first I would ask you to go back and write out in far more detail what you want to accomplish. How long you have for each phase and what ability do you need at each step. It maybe better to spend a little more in the beginning rather than shorty learning you need to swap out an expensive part(s). For example. If your end plan is to do obstacle avoidance by the end of year one. Then a simple controller simply won't get you there. But a good controller like the APM2 with the ability to add additional sensors later, would be a better cost effective choice in that first year. A $100 controller you throw away in 9 months doesn't make sense. But if the same goal is over 4 years and the possibly of more funds next year, then perhaps it does make sense. But I think you need to answer some of those questions first.
This is as much help as I can be atm. There are better people here than I. But I believe, you need to detail your plan and time lines, along with your budget a little better, if you hope to get where you want to be and before you worry about motors and props.
Hi Terry and thank you for replying to my post.
You've raised a few questions that I hadn't even thought about in too much detail to be honest and I've spent some time since reading your reply mulling them over.
Basically, for the first 6 months, all I will be doing is learning about the system as it stands.
I'll be looking at how the software works, how the software interacts with the hardware, test flying and generally experimenting with the existing system before I dare make any modifications or alterations to it.
From there, I hope to start investigating / experimenting with situational and environmental awareness.
I want to look at giving the craft certain decision making abilitites based on the information it knows about its surroundings and environmental conditions.
Later in my degree, we are going to be looking at Fuzzy Logic which will prove to be a useful tool for furthering my investigations / experiments into awareness and decision making.
My time frame for this and any other areas I explore will be over the next 2 - 3 years so I'm in no hurry but you are quite correct in your observation that I would much rather be developing flight control software than investing most of my time trying to "re-invent the wheel".
So, based on some of the questions you raised and observations you made, I've been and re-evaluated my goals in more detail and I've also assessed my needs better and I think I've found something that will keep me happy for the next year or two without the need for many upgrades / expansion in between times.
I've found a blog online with a chap who built a quadcopter on quite a tight budget and it performed really well barring a few intial teething problems which he resolved so I hope to follow his design but instead of using the KK controller he was using, I'm going to be fitting the APM V2 - when its available. =]
You were also quite correct about my budget so with permission from my good lady, I've managed to up my limit to £400 which allowed me to go with the APM which is what I really wanted. I totally agree, it's a far superior board and there's no point scrimping where the brain of the craft is concenred.
So far, I've been out finding everything I need and cost watching and I've managed, on paper at least, to get everything I need from RC Transmitter to Propellars for just under £400.
I've had to go here, there and everywhere to find the best parts for the least amount of money and while that does mean waiting on certain parts that aren't in stock yet, I will eventually have a platform I know should work based on the fact that I've seen the design working on video.
Thanks again Terry,
A bit of clarity and an outside perspective always helps =]
It wasn't much Mike, but it was my intention and hope that in what I wrote it would kick start you into examining your goals. In as much as you are aware of details now so early in your plans. It isn't possible at the beginning of a project to be aware of all the variables or pitfalls that will occur but establishing a good grounding as possible is important. One thing I have learned, is to try to keep a plan B for each critical junction of your research. Try not to box yourself into only one path. It's not so bad if you do so in software, but when hardware is involved it becomes more than just lost time and can be a double budget killer that halts everything. That being said, what you plan on will take a lot of horsepower The good news is the apm2 is one of the most powerful for the price out there right now and should take you far down the road in your research. Hopefully you are one of the new bright sparks that will advance new and efficient algorithms, methods and protocols.
Good luck and all the best Mike. As this board and it's predecessor have a very large user base, is open source, so linked to many other efforts and it's your best chance to discover the many experienced people around here in the forums, blogs and outside in related forum groups for advice, like rcgroups. Hopefully I'll be here over the long term, to see you around as you report back your successes. ;)