PDA

View Full Version : Programmable Microcontroller Modules



LinuxGuy
05-21-2008, 07:01 PM
<Crossposted for further discussion>

Re: The Parallax (http://www.parallax.com) Propeller


That's super reasonable for that much processor. Looks interesting. I love microcontrollers. ;)
The 40 pi DIP Propeller is going for $12.95 or so now. That's real hard to beat for a module of any kind. I have one of them around here somewhere. I need to to find it. :) The only bad thing about the Propeller, from my perspective, is it the development software requires Windows.

I'm done working with modules that cost $59.95 or more each, especially ones with flaky development software. I have two remaining Basic Micro (http://www.basicmicro.com) modules here (a Basic Atom, and an Atom PRO), and when those are toasted I will not buy more. Both cost $59.95. The software for the Atom PRO sucks, so I don't do anything with it. The software for the Basic Atom is solid though. Both require Windows to develop for, as does the Propeller. How hard could it be to create Linux/UNIX friendly development IDEs? Maybe it's more difficult than it seems to me.

It's C programmable and Linux development friendly for me from now on, as much as possible. I think Propeller could be an exception though, and it's very inexpensive for what you get. It wouldn't hurt as much to blow a Propeller as it would a module that costs $59.95 or more.

8-Dale

Wingzero01w
05-21-2008, 09:28 PM
The only thing about propeller is it requires knowledge of both low level assembler/high level spin, after you get through that, you can take advantage of its features.

Im still trying to get ahold of the chips you mentioned to me before on another thread to start embedded computing. At the same time im looking for a powerful chip for a board i want to make. Im looking for a fast (200+ Mhz) low pin (44-300) with decent memory and usb/i2c and all those types of interfaces.

ooops
05-21-2008, 09:32 PM
Robotguy, thank you that is handy info, for a newbe like me.
Hey look my first post ... I have been following for a few weeks now, and have read all the recent threads. I look forward to leaning on you and the rest for help in coming weeks. Keep the great info coming!

LinuxGuy
05-21-2008, 09:52 PM
The only thing about propeller is it requires knowledge of both low level assembler/high level spin, after you get through that, you can take advantage of its features.
I've looked seriously at the Propeller, but not recently. I don't recall there ever being a requirement to do anything with assembly language. It looked to me like everything could be done in SPIN, with the option always being there to also code in assembly.


Im still trying to get ahold of the chips you mentioned to me before on another thread to start embedded computing.
It's not as difficult as you seem to be making it. Go to the website, click samples, register if you aren't already registered, pick samples and create an order, change quantities to the maximum for each part, checkout. Pick a technology sounding business name when you register. Make stuff up where you have to - they really don't seem to care what you put in there.


At the same time im looking for a powerful chip for a board i want to make. Im looking for a fast (200+ Mhz) low pin (44-300) with decent memory and usb/i2c and all those types of interfaces.
Don't get ahead of yourself now. Learn some basic microcontroller programming and do some simpler projects first. Blink some LEDs and then sequence LEDs and then PWM the LEDs. Move on from there and try some simple motor control. Then try sensor processing for something like a PING. Learn how to work with I2C and SPI. These are all things you will eventually need to learn for more advanced projects.

What is the project you want to do with the fast processor? You are probably looking at an ARM9 class or better processor to get that speed range. Maybe an AVR32 will fit. You may not actually need that much raw speed depending on your application.

BTW, I have spent over a year so far working with PICs/dsPICs before moving into more advanced stuff like ARM7 and ARM9. Everything I have learned, I am building on now.

8-Dale

Wingzero01w
05-21-2008, 10:22 PM
Thanks Robotguy,

Lol i always thought i needed a buisness to get the chips, i didn't know that they didn't care about it.

Well from what I've read on parallax forums, you can do most programming on the Spin language but for some areas there might be need for assembly i think this is for really advanced applications.

And the application i want to do with the processor is AI, i know theres stuff like the Pico-itx and such... but thats boring... id rather make my own board and gain from the learning experience.

LinuxGuy
05-21-2008, 10:31 PM
And the application i want to do with the processor is AI, i know theres stuff like the Pico-itx and such... but thats boring... id rather make my own board and gain from the learning experience.
I am very interested in AI also, and am looking at AI programming languages now. We can continue to discuss that in the other thread. :veryhappy: I favor the distributed approach to robotics.

8-Dale

Wingzero01w
05-21-2008, 10:41 PM
I am very interested in AI also, and am looking at AI programming languages now. We can continue to discuss that in the other thread. :veryhappy: I favor the distributed approach to robotics.

8-Dale

After i get the funds, i plan on buying the pheonix, putting on some 985MG servos, learning the gaits and how to walk that and then dissasemble it and make the quadropod ive been dieing to make. For both the pheonix and the quadropod, im going to make a simple AI for enviorment reactions, on the humanoid however- i want to make it express emotions and have more of a human reaction.

Back to this thread, ive also been wondering... if you can make two microcontrollers act like two cores of one big processor- like a dual core cpu on a computer. Im saying, not have them both take care of different tasks, but take care of the same task together.

LinuxGuy
05-21-2008, 11:33 PM
Back to this thread, ive also been wondering... if you can make two microcontrollers act like two cores of one big processor- like a dual core cpu on a computer. Im saying, not have them both take care of different tasks, but take care of the same task together.
Yes, you can do this. :veryhappy: It's commonly called clustering. There are several Linux packages that allow doing this, such as Beowulf and others. There is an Open Source project called OpenMosix, aimed at making this possible at the kernel level. Through creative scheduling, it might even be possible to do this with some microcontrollers that do not run Linux.

I also have two NXP LPC-2148 boards - a header board (basically just the CPU, USB, and some headers to bring out the pins for easy access) and a proto board made by Olimex. I've had FreeRTOS (http://qqq.freertos.org) built and running on the proto board, but never did customize it for the hardware. I really like the NXP chips for ARM7, but I am also getting real interested in the Luminary Microsystems (http://www.luminarymicro.com) microcontrollers. I particularly like their LM3S6965 (http://www.luminarymicro.com/products/lm3s6965.html) because it has two hardware quadrature encoder modules (for wheel encoders). I would love to pair one of these with Hammer (http://www.tincantools.com/product.php?productid=16133&cat=0&page=1&featured). The last chipw (as of now) that I am really interested in are the Freescale (http://www.freescale.com) i.MX31 (473 ball BGA) and Atmel (http://www.atmel.com) AVR32 AP7001 (208 pin QFP).

BTW, if you've never been to the Spark Fun Electronics (http://www.sparkfun.com) site, go there soon. You want gadgets, they have gadgets.. :veryhappy::veryhappy:

8-Dale

Wingzero01w
05-21-2008, 11:45 PM
Yes, you can do this. :veryhappy: It's commonly called clustering. There are several Linux packages that allow doing this, such as Beowulf and others. There is an Open Source project called OpenMosix, aimed at making this possible at the kernel level. Through creative scheduling, it might even be possible to do this with some microcontrollers that do not run Linux.

I also have two NXP LPC-2148 boards - a header board (basically just the CPU, USB, and some headers to bring out the pins for easy access) and a proto board made by Olimex. I've had FreeRTOS (http://qqq.freertos.org) built and running on the proto board, but never did customize it for the hardware. I really like the NXP chips for ARM7, but I am also getting real interested in the Luminary Microsystems (http://www.luminarymicro.com) microcontrollers. I particularly like their LM3S6965 (http://www.luminarymicro.com/products/lm3s6965.html) because it has two hardware quadrature encoder modules (for wheel encoders). I would love to pair one of these with Hammer (http://www.tincantools.com/product.php?productid=16133&cat=0&page=1&featured). The last chipw (as of now) that I am really interested in are the Freescale (http://www.freescale.com) i.MX31 (473 ball BGA) and Atmel (http://www.atmel.com) AVR32 AP7001 (208 pin QFP).

BTW, if you've never been to the Spark Fun Electronics (http://www.sparkfun.com) site, go there soon. You want gadgets, they have gadgets.. :veryhappy::veryhappy:

8-Dale

Awesome i really gotta learn how to do clustering... i got nice board ideas for that.

Im not completely sure but i think luminary micro made the chip for the iphone... might be wrong, but its extremely nice as a controller. With all these chips on the market it really is hard to choose the right one for an application. Ive always wondered though, whats the difference between programming microcontrollers from luminary, atmel, microchip, etc. and programming something like the intel atom?

O yes i have been to sparkfun and love the site, lots of stuff to look at that can help with whatever application. They even offer lots of nice services.

Adrenalynn
05-22-2008, 02:00 AM
You gotta learn some basics first and stop getting ahead of yourself. I appreciate your enthusiasm, but at some point you need to face reality and start where we all start - at the start.

Modern assemblers are trivially easy compared to yesteryear, stop fretting about it. If you can learn basic, you can learn modern assemblers.

Start with the PC. Use simulators and emulators. Learn the basics or you're just throwing away your time and your money. If you don't build on a good foundation anything you build will crumble into rubble.

LinuxGuy
05-22-2008, 03:47 PM
You gotta learn some basics first and stop getting ahead of yourself. I appreciate your enthusiasm, but at some point you need to face reality and start where we all start - at the start.

<snip>

Start with the PC. Use simulators and emulators. Learn the basics or you're just throwing away your time and your money. If you don't build on a good foundation anything you build will crumble into rubble.
Well said, and I agree completely. Even I started out with a very basic robot kit. That's how W.A.L.T.E.R. came about, and now look where I can take him. :veryhappy: He's been almost completely rebuilt from that kit, and the only original part now is the front caster. :happy:

Yes, new folks to robotics need to start out basic and work up. I can't emphasize enough how important this is. Everything learned is built on for the next stages - even if the type of robot is completely different. My next major robot project will be a hybrid of some form, maybe even a W.A.L.T.E.R. hybrid. :veryhappy::veryhappy:

8-Dale