General Misc > Misc

Im back, heavy loaded, but no firearm...

(1/2) > >>

Waaaaaaazzup!!  :)

I have been disconnected from the world of robot development for 5 years or so.
Now om back and i want to build again.

I have a 3d printer and can see that arduino had comed to the market.
I also have a pic start plus programmer and a development board.
Has the pic controllors been replaced by arduino?

Is C programming still the best software code?

What will the future software code launghes be? Will C or C++ still be popular?

I want to build a 2 legged standup robot, but try to see of the mechanical part can be better than only using servoes. Anyone who has ideas? How have they build the joints for the advanced robots, ex the one honda did make years ago?

     I would think that to some degree the PIC controllers have been replaced. With the primary disadvantages with the Arduino boards being: pricy and lacking power. The price can be corrected with building the Arduino board yourself, building the Arduino UNO cost approximately $15 versus $30.

Building an Arduino:

The power/speed disadvantages can be solved by using the Arduino DUE, I haven't had the DUE for very long, but so far I really like it. The DUE is priced at $70, rather pricy, but you could look into the Raspberry Pi (I prefer tau :)  I have no experience with the  Raspberry Pi, but most people seem to like the board. Plus it is very comparable to the Arduino DUE, I don't know about the support/community around it thought (it's hard to beat out the Arduino in this aspect)

Arduino DUE:

Raspberry Pi

In my opinion C++ is more versatile due to it being a multi-paradigm, that is being object oriented as well as being procedural. What will be popular in the future is beyond me, but as a educated guess I would say C++ will get more popular (in the near future).

Taking as much mass off the servos is the best way, how to go about this exactly is a tricky matter. I don't know the size of your humanoid robot, but if it is a considerable size I would look into Dynamixels (if the budget allows)

Dynamixel servos:

The Due is a dead product IMO. It is much slower than the Raspberry Pi (by a factor of 10 or so) and it has much less RAM (by a factor of a few thousand.) Neither the Due or the RPi are 5V compatible, though.

The Arduino Uno is pretty nice for tinkering. If you want that same CPU on your own (Atmega 328p,) it's only $2.50 in quantity dozens, and it can run just fine with just supply voltage and a de-couping capacitor (which costs maybe a quarter) and perhaps a reset pull-up resistor for good measure. You'll need a USB programmer, rather than a bootloader, to program it, but that's pretty cheap. And you'll need a 6-pin ICSP header to actually program it, unless you have some socket that's already wired to plug it into. It will run on the built-in RC oscillator at 1 MHz or 8 MHz. If you want up to 16 MHz (the Arduino Uno speed) or 20 MHz (the rated speed) you need a crystal (50 cents) and a couple of small capacitors (a handful of cents.)

Breadboard, CPU, capacitors, crystal, programming header, and resistor is probably less than $6, somewhat depending on the cost of the breadboard. And the programmer is $20, and a one-time investment.

Finally, the reason the AVR has such a fan following is that it is very easy to program. The programming model makes sense, and peripherals are very straightforward. avr-gcc is a great compiler, and avr-libc is a reasonably C library (but, really, in an MCU, you don't want to be using too much of the C library...) avrdude makes uploading/programming super easy. And here's the secret: The Arduino IDE actually uses avr-gcc, avr-libc and avrdude under the covers. It just puts a newbie-friendly front on those tools.

I like the DUE just because of the support/community around Arduino, but really I should buy and start working with the Raspberry Pi.

Thanks so much for your answers! Really good.

I see were We're going.. my old pic programmar has had its glory..

And thanks for all the links!

Personally om much møre into the mechanical aspect of this world.
So the electronic and software is gonna be a big challenge I see.

I have decided to start making one of the arms. Concentrating on building the joints for the shoulder and hands. This scale is probably gonna be 1:1 at first.

The movement of the fingers and wrist might be either air controlled or hydraulic or using wire and some sort of.control. As a first test I'm gonna start using plastic needels as actuators.

I would still love to discuss other solutions.


[0] Message Index

[#] Next page

Go to full version