Society of Robots - Robot Forum
General Misc => Misc => Topic started by: blackbeard on February 07, 2011, 02:15:30 PM
-
how come nobody ever wants to make their own hardware? everything i read about the arduino involves some sort of mass produced super expensive shield. what use is hardware development if it holds together like lego!
-
how come nobody ever wants to make their own hardware? everything i read about the arduino involves some sort of mass produced super expensive shield. what use is hardware development if it holds together like lego!
Nobody forces you to use Arduino. Arduino is essentially a repackaged ATmega CPU, and these are extremely easy to use, program in C, and interface to other components, without the need to pay 1000% premiums to anyone.
The reasons why people use Arduino are because it has a vibrant community; because it's well-marketed, so the alternatives may be less obvious in comparison; and because some folks are unwilling to read MCU specs or design their own PCBs. It's an interesting phenomenon, but it's probably not very productive to get too worked up about it ;-)
/mz
-
I think the reason why more people don't design their own PCBs is that their robots aren't designed by a business. The robot is a project, not a product. They aren't designed to be mass-produced so cost savings isn't as important (other than how much money the hobbyist can spare). Minimization and form-factor aren't as important either as there is no one to please but themselves and no other product to compete with. Also, of course, is the learning curve and hassle that comes with designing your own PCB.
Personally, I agree with you - build quality is important to me and I like to build things that look professional while still being custom. I'm trying to design my own PCB right now that will be produced by a factory that fits my robot like a glove: small form-factor (SMDs), just enough connections for the components on the robot. It's tough, though!
-
you don't need to produce your own boards really. you can use protoboard like i do. in fact most of my projects use protoboard with a socket for the atmel328 rather then the arduino it's self so i can pop it out later and use it in other projects or simply leave it in if and buy a new $5 chip.
-
If you don't want to design & etch, perfboards aren't exactly rocket science. They look pretty OK, though obviously not as nice as a silkscreened PCB.
-
If you don't want to design & etch, perfboards aren't exactly rocket science. They look pretty OK, though obviously not as nice as a silkscreened PCB.
oh i've made my own pcb before (without a drill this is harder then you think) and i'd do it often except i like to modify my circuit as i go most of the time. i guess i'm one of the few people who prototypes with a soldering iron rather then a breadboard or protoshield
-
Hi,
how come nobody ever wants to make their own hardware? everything i read about the arduino involves some sort of mass produced super expensive shield. what use is hardware development if it holds together like lego!
Hey, I wouldn't buy a shield if it couldn't stop a sword... Call me oldfashioned, but I hate it when they invent pop-silly names for things that are already well named - like extension modules.
Who was the moron that "invented" this?
Can we put out a contract on him/her, or at least get his/her arms broken in several places? ;)
A name should be descriptive.
A shield has had a certain purpose for at least a couple of thousand years.
Only insane people would go into a sword fight with an extension module.
'nuff said.
-
Who doesn't love the smell of fresh and hot ferric chloride in the morning... Gives nice colors also...
Look... It takes some time to actually start designing your own circuits and do some more complex stuff...
Unfortunately, most boards in the market are "SMD part to board" type... not much ingenuity...
Like Sparkfun modules for example...
It's always nice building you stuff... And Soerer... Unless you are in an urge to go to war.... Customizing - building your own shield can be a live saver as well as self-made trap...
Treat everything in life with love and compassion and you shall always have nice results...
Be patient people... Rush things and they may work but can get no satisfaction... :-p
There's nothing better than seeing your labor working and being good looking... ;-)
-
Hey, I like modular stuff, it's dirt cheap when you think about it, I could spend three weeks building some board soldering it together with cpu, timers, rs232 the likes only to have some glitch y section start to fart around and spend another 3 weeks trying to find the problem on the board then replacing the parts again. OR unplug module, replace with new module ,five minutes get back to writing neat programs for my bot.
I know it's not that simple to some but if my current bot craps out electronically it could literally take weeks to find the problem if it wasn't for modules and program diagnostics , each type of module can be tested through programming to find the errors. Let the machine do the work of diagnostics it's more fun.
-
Hey, I like modular stuff, it's dirt cheap when you think about it, I could spend three weeks building some board soldering it together with cpu, timers, rs232 [...]
You realize that ATmega already has pretty much everything on-chip?;-)
-
What like wireless rs232, sensors etc.
-
What like wireless rs232, sensors etc.
I'm saying I like the shields, the modules etc, I'm too old for this soldering crap, I can read a manual that says do this , do that and when I put it all together it works, I love it.
I'm not familiar with Arduno and Atmega's, I'm using an Atom28 Pro on an Atom Bot Board II and an SSC32, I find it very easy to program. My bot has 3 X-Bee wireless rs232 modules plus various sensors, I would have never attempted to solder this together I would probably have died of old age by now.
-
What like wireless rs232, sensors etc.
Not quite, but you listed
[...] with cpu, timers, rs232 the likes [...]
All of which are, in some form, on the average ATMega328 microcontroller. Admittedly the UART is TTL and not RS232, but that's just a quick level-shift away.
I believe that Arduino "shields" differ from generic extension models in price and form-factor: the ability to build Arduino based projects like Legos appears to permit a significantly higher price.
I will freely admit that sensors and wireless links aren't included in a microcontroller - it allows for modularity. If I wanted to use XBee, I would buy an XBee chip - but I'm pretty sure I still wouldn't buy the Arduino shield. Extension module's aren't quite "dirt cheap", but they fit the cost-benefit curve nicely in some cases:
Examples
Stereo Vision:
Shield-style: Buy a $5000 commercial stereo vision system - easy and guaranteed to work
Module-style: Buy two $50.00 cameras and set up the system yourself
Hardware-design style: Build a custom PCB and install two $2.50 CCD or CMOS cameras on it
Servo Controllers
Shield-style: Buy a shield that breaks out unregulated voltage, ground, and signal in servo-style headers, and put the servos on that
Module-style: Buy a piece of stripboard and make your own power buses
Hardware-design style: Build a custom PCB with servo headers
All have their own advantages - I personally waver between modules and making my own hardware, but will buy a "shield" (When I say shield I really mean overly-expensive proprietary module that is comparatively easy to use) if it's absolutely necessary.
P.S.: You could just edit your own post instead of quoting it - it confused me a bit to see that you responded to yourself.
-
What like wireless rs232, sensors etc.
Not quite, but you listed
[...] with cpu, timers, rs232 the likes [...]
All of which are, in some form, on the average ATMega328 microcontroller. Admittedly the UART is TTL and not RS232, but that's just a quick level-shift away.
I believe that Arduino "shields" differ from generic extension models in price and form-factor: the ability to build Arduino based projects like Legos appears to permit a significantly higher price.
Examples
Stereo Vision:
Shield-style: Buy a $5000 commercial stereo vision system - easy and guaranteed to work
Module-style: Buy two $50.00 cameras and set up the system yourself
Hardware-design style: Build a custom PCB and install two $2.50 CCD or CMOS cameras on it
Servo Controllers
Shield-style: Buy a shield that breaks out unregulated voltage, ground, and signal in servo-style headers, and put the servos on that
Module-style: Buy a piece of stripboard and make your own power buses
Hardware-design style: Build a custom PCB with servo headers
The biggest problem is "who knows to make all these work together...
I'm trying nowadays building my own vision system based of some roughly 30$ camera's each...
Of course I'll start with a single camera I'm not getting funny around it...
But the cost for a single camera set up as hardware saying is about 60 - 70 $ Maximum...
The biggest problem is elsewhere...
I must also have it working... Meaning the biggest trouble would be coding.... Coding it to work, coding it to be compatible with other machines coding it for anything else...
Finally, If I wanted to place a price tag and go to mass production, I would say that 15 - 20 $ dollars a piece for my work would be could enough, as a vision system isn't that easy to make, and not so common...
Ending price would be anywhere between 80 - 100 $ which is a nice price, as you can see any other vision module on the market has such a price...
About servo controllers... (same as LCD serial modules)
Well... They are over priced... The program they are running is basic stuff in programming...
I'll code one for you to see... ;-)
-
how come nobody ever wants to make their own hardware? everything i read about the arduino involves some sort of mass produced super expensive shield. what use is hardware development if it holds together like lego!
The same reason people don't injection mold their own lego pieces.
Not everyone using Arduino wants to become an EE. They have an idea and want to implement it the fastest way possible. Artists, computer-scientists, etc don't want to learn or delv into hardware design, they just want something that works. Arduino people foresaw this niche and went after it.
-
The Arduino was designed to be ultra cheap, not functional. There are no power buses built in to the Arduino, so you can't just plug anything but signal lines into it. As such, to use one, you are forced to either use A) a perf board or B) a expensive shield. Try plugging something as simple as a servo into an Arduino and you'll quickly see what I mean.
Sometimes I wonder if this was intentional, to get you to buy the Arduino because it looks very cheap at first, but then they sell you expensive add-ons later after you're locked in. Then you need to buy a wiring kit, a protoboard, male headers, and two batteries to run a single servo.
That said, the Axon (warning: self serving plug) doesn't require external addons. You don't need shields, perf boards, or multiple power sources. It's a single purchase, no hidden costs. /end of self serving plug :P
-
My take on this is to have it both ways. I design my own boards with either a 328 or 1280/2560, add an FTDI chip and I can use the Arduino environment (not the IDE that's crap, I use Netbeans) and get most stuff running pretty quick. There is always the option to do direct hardware manipulation or (with an ISP header) write in assembler using AVR Studio should I feel the need.
So in a nutshell I like having the bootloader and libraries, but seldom use existing hardware except for proof of concept.
______
Rob
-
I fought with a co worker for days about the arduino. I said buy an axon they do more stuff. But he is lacking in the code department and wanted something easy. I'm still trying to get him to upgrade.
On the other hand, My old electronics professor just bought a ton of them for her students. We talked her into those. They are easy enough that freshmen can start getting involved and do cool stuff. First micro-controller I saw in college was as a senior and that was the Z8 Encore. It limited our senior design project because the learning curve for the Z8 was HUGE. AVR is better but not by much. Last year the freshmen learned to program in basic on a computer for the same class they are now designing functional circuits and program arduinos.
Both have their place but if I was investing in something I will use a thousand times then I go with the most versatile if its simplicity I need then arduino works.
That said the shield idea always confused me and I don't see the point but modules have their place and are useful. Now where did I put that OSCAR project....
-
A modular prefabricated system is easier for the main audience of the arduino, artists, hobbyists, newbies...
you're welcome to make your own arduino boards though.