hi,
another arduino noob here and loving my new toy. i'm still trying to learn and play with the arduino programming language. yup, still working on blinking leds and ldrs.
first project that i've been planning for is to build a small diff drive robot.
i have a quick, and i believe simple, question. my understanding is the arduino is a development platform. why am i seeing a lot of folks using their exact same arduino's, shields and all, on their robots? do these folks buy new arduino's on their next robot/project?
or maybe, their robots are really never done and they will keep adding to it?
please correct me, the workflow as i understand is...
develop your program in the arduino demulawhatever (i know it's 2009) with all the sensors, servos, etc. hooked up on a breadboard.
test, test, test your program
test, test, test some more...
when done, take the chip out of the arduino, insert it into a bare minimum PCB - bare as in only the atmega on a socket, the resonator, hookup or the actual +5V regulator, headers for the servos and sensors, etc. - live test and if all goes well, "go live".
if there is a problem or if i need to add features to the robot, take the chip out, into the arduino again then goto 1.
on to a new project, get a new atmega328 chip with bootloader.
That does seem to be the general idea, if you're planning to build a long-term project. New ATMega328s with Bootloaders are only about $5 USD, so it's relatively cheap ;).
Yes, it is a development platform. Once you've developed the robot to the extent you want, it's no longer a challenge. So, you take it apart, and move on to something different.
If you were designing and building something that had a real purpose, you would do just as you outline.
For many "real" projects where your time is not for free, especially when there is no mass production, just a single unique product, it is faster and cheaper to leave the Arduino board in the product and buy a new one for the next project.
No time and cost to design and test a PCB, the system can be updated without having to move the processor. If it is not hobby work (this is not criticism or negative) and there is a price tag on your time then the cost of the Arduino board is paid for in one hour.
or maybe, their robots are really never done and they will keep adding to it?
This is probably closest to the truth. More charitably, many hobby robots are "experiments" and are never really MEANT to be "done."
when done, take the chip out of the arduino, insert it into a bare minimum PCB - bare as in only the atmega on a socket, the resonator, hookup or the actual +5V regulator, headers for the servos and sensors, etc. - live test and if all goes well, "go live".
Keep in mind that having a "bare minimum" PCB made is likely to cost more than the arduino module. If the "development platform" works, and is cheaper than "going live", AND you can take the pieces off and reuse them for something else if you need too, what was the advantage of making a "pure" finalized "product" ?
I hate admitting that projects are "done", so the Arduino stays in
More practically, if I want a USB port, easy reprogramming, voltage regulator, and reset button, it may be faster and cheaper to keep the Arduino permanently emplaced than sort out a one-off standalone system. For me, the major reason to replace the Arduino in a unique device would be case space, if it were at a premium.
i guess i knew my answer all along but just wanted to hear from other folks. this sort of hobby is most of the time more like an ongoing experiment which is why it usually stays in.
but if i were to build say a robot toy for my kids who will be happy with something that just walks around avoiding objects or following them, it is something that has an "end" (or is it?). i'd feel better running it on say a $12 minimal http://moderndevice.com/RBBB_revB or something similar. something similar.
I know that I go an on on about the Boarduino, but I find that it has a form factor that lends itself to a more permanent non-development solution.
It's real easy to treat it like a "stamp" and just build a PCB around it... using various shields for development and moving the "shield" designs to a motherboard you insert the Boarduino "into" to get you away from multiple stacking.
as I have always been curious as to all these different 'models' of the Arduino boards..mega, nano, pro..mini..etc...
Its more or less the 'boards' and the extra hardware/components that make the model's different?
or are the actual CHIP SETS different? like the mega has 6 hardware interrupts to use..and the others only 2 I think?
so would I need to get the MEGA ($66+) to develop with..
then once I test..etc..etc take my chip (or get same exact chip type).. flash with my Arduino boot/code..etc.. and pop that chip intot he final design of my choice? (that mimics the functionality of the breadboard/development layout?)
This comment has peaked some interest for me.. can anyone elaborate?
The RBBB is the way to go. I bought an arduino NG board years ago, back when it still used atmega8. Since then I've bought nearly a dozen RBBBs, they are great for small enclosures and things that fly like my helicopter.
The first three chips are pretty simular to what they can do (except for memory space for your programs) Also the first tree chips are availlable in dip socket which is makes it easy to handle in d.i.y. projects.
The 1280 chip which is used on the mega has lot's more IO pins, so you can connect a lot more stuf.
Basicly you can develop on every model as long as it fits the memory.