This blog follows progress on a small scratch-built VTOL platform. Sci-Fi styling and modeling are just as important to me as functional operation, even if it compromises the design. I'm doing this because (while) I feel like it, no other reason. See first post (intro) for more...
Friday 6 August 2010
Ardunio or PIC - Action Priority Matrix
I remembered this:
I don't want the tool to be a major project, unless it needs to be. Arduino's looking good!
All from scratch will always meet your needs better, but will also probably cost you way more then you wanted to spend in re-designs (I've noticed you've been having all that fun to).
My UAV project is beginning to come together and on the controller front I've built a custom STM32 board - it will sink most PIC's and put an Arduino in the weeds on a performance front, but my quest for perfection has also now taken it to "3rd spin coming soon" [Just to tidy up some silkscreen].
I've also built a IMU board that fits perfectly on the footprint of my controller (a benefit of totally custom). But in 3 days I realised that you should avoid coupling your main controller with orientation calculations if you can avoid it.....you can try and isolate the code blocks, but at the end of the day your end solution is very inflexible. So I'm in the process of redesigning that whole approach to use a custom "position data concentrator".
The very best thing about the latter solution of decoupling your main controller from how it gets to know where in the world it is, is that you can then feed this data in from any source (i.e. A simulator).........and yes I've been very busy developing a Software/Hardware In the Loop simulation platform for X-Plane.
In the end how good do you want your results to be Vs how long do you have.......custom will drive better end results, but you'll be at it for much longer (and sink more cash [and time] then you ever thought on day one). But if you want it to be right (and very re-usable) at the finish, then I'd say you want to take a custom approach - but always keep in mind "but how I could I use my tech. later" at every decision point you make.
Hi Phill, think I saw you in Milton the other day... Totally agree with you. I think I'm going to go for an Arduino just for quick hacks (I have trouble getting anything done as it is).
Did ERIX run out of steam (like most of my projects do)? What hardware architecture are you planning to use for your UAV, quad-rotor? I couldn't find much info on your website.
ERIX looks really good and sits up on "jacks" on my desk for much of its life, I really should take some photo's, its far away from when I stopped writing about it (I'm really not good at documenting this stuff). It's now my development platform for all new "technology" - so it's eternally evolving....and that's basically its goal in life now. I reached a point on the ground vehicle stuff where I realised that at sub-car scale there's no real automous application.....but as a development platform they work really well (i.e. if you want to improve and improve a motor driver or a controller board etc. they're great proving grounds)
My UAV is aimed at a fixed wing. One lesson I've took from ERIX was that you can sink a lot of effort into custom mechanical stuff and have nothing but a lump - then you need to find some more magic time to make it do something....So the UAV is going the opposite way, I've spent many hours developing the simulation platform and building the controller & writing board test and control system software (and much, much less time worring about the chassis).
The plane itself is the next significant purchase, so there should be lots going on over the next couple of months as it's all beginning to come together now - where as for the last 5 months its been a lot of PCB's and wires and stuff everywhere on my desk.
One of the other goals of course was to come out of it with a standard controller platform for the "everything else" I might do later. On that front the main controller ticks the "Objective Achieved" box :-)
Hey Paul hope all is well,
ReplyDeleteAll from scratch will always meet your needs better, but will also probably cost you way more then you wanted to spend in re-designs (I've noticed you've been having all that fun to).
My UAV project is beginning to come together and on the controller front I've built a custom STM32 board - it will sink most PIC's and put an Arduino in the weeds on a performance front, but my quest for perfection has also now taken it to "3rd spin coming soon" [Just to tidy up some silkscreen].
I've also built a IMU board that fits perfectly on the footprint of my controller (a benefit of totally custom). But in 3 days I realised that you should avoid coupling your main controller with orientation calculations if you can avoid it.....you can try and isolate the code blocks, but at the end of the day your end solution is very inflexible. So I'm in the process of redesigning that whole approach to use a custom "position data concentrator".
The very best thing about the latter solution of decoupling your main controller from how it gets to know where in the world it is, is that you can then feed this data in from any source (i.e. A simulator).........and yes I've been very busy developing a Software/Hardware In the Loop simulation platform for X-Plane.
In the end how good do you want your results to be Vs how long do you have.......custom will drive better end results, but you'll be at it for much longer (and sink more cash [and time] then you ever thought on day one). But if you want it to be right (and very re-usable) at the finish, then I'd say you want to take a custom approach - but always keep in mind "but how I could I use my tech. later" at every decision point you make.
www.qudon.com
Hi Phill, think I saw you in Milton the other day...
ReplyDeleteTotally agree with you. I think I'm going to go for an Arduino just for quick hacks (I have trouble getting anything done as it is).
Did ERIX run out of steam (like most of my projects do)? What hardware architecture are you planning to use for your UAV, quad-rotor? I couldn't find much info on your website.
ERIX looks really good and sits up on "jacks" on my desk for much of its life, I really should take some photo's, its far away from when I stopped writing about it (I'm really not good at documenting this stuff). It's now my development platform for all new "technology" - so it's eternally evolving....and that's basically its goal in life now. I reached a point on the ground vehicle stuff where I realised that at sub-car scale there's no real automous application.....but as a development platform they work really well (i.e. if you want to improve and improve a motor driver or a controller board etc. they're great proving grounds)
ReplyDeleteMy UAV is aimed at a fixed wing. One lesson I've took from ERIX was that you can sink a lot of effort into custom mechanical stuff and have nothing but a lump - then you need to find some more magic time to make it do something....So the UAV is going the opposite way, I've spent many hours developing the simulation platform and building the controller & writing board test and control system software (and much, much less time worring about the chassis).
The plane itself is the next significant purchase, so there should be lots going on over the next couple of months as it's all beginning to come together now - where as for the last 5 months its been a lot of PCB's and wires and stuff everywhere on my desk.
One of the other goals of course was to come out of it with a standard controller platform for the "everything else" I might do later. On that front the main controller ticks the "Objective Achieved" box :-)