Robots building using chipKIT boards

Created Sat, 23 Mar 2013 22:56:55 +0000 by nightboy

nightboy

Sat, 23 Mar 2013 22:56:55 +0000

I am looking for a microcontroller board that can control many motors and sensors. Arduino boards might be the best in doing this task but they do not have a fast CPU and RAM like chipKIT boards (better in programming ). Unfortunately, I do not have that much experience in electrical circuits and I could not find that many resources for chipKIT boards that explain how to use them in building robots. So, my question is: what is the best chipKIT boards for building robots with many digital sensors and motors (servo & DC) ?

Note: by "many" I mean at least 10. so, at least 10 sensors and 10 motors.

guymc

Mon, 25 Mar 2013 23:42:15 +0000

Several robot shields and kits are available to work with Arduino and chipKIT development systems. You might want to take a look at these:

http://www.seeedstudio.com/depot/shield-bot-p-1380.html

https://store.wayneandlayne.com/products/bricktronics-shield-kit.html

http://www.abotics.com/open_robot.htm

Some kits from Digilent are based on Cerebot boards, which are software-compatible with the chipKIT platform:

http://digilentinc.com/Products/Catalog.cfm?NavPath=2,403&Cat=11

There's a lot more out there; start with a web search and dig in!

Cheers

nightboy

Tue, 26 Mar 2013 05:56:03 +0000

I have looked at them but non of them could be used in controlling medium to large size robots (Many sensors and motors). For Example: 6 10A DC motors could not be driven by any shields or Pmods. On the other hand, Arduino boards have plenty of shields that can drive real big motors(40A).

I really do not get it. How could it be possible to build big robots using chipKIT boards ?

majenko

Tue, 26 Mar 2013 10:00:11 +0000

By not relying on other people's work and designing your own circuits?

To drive "real big motors" (by the way, 40A is not big. 400A is big.) you just need to connect them up using the right components - such as a MOSFET that can cope with the current you will be driving.

For those kind of power levels it is often better to have the control circuitry close to the motor to keep the high current cables short and just route small, flexible, cheap signal wires to the individual motor's controller circuit to control the motor. You don't want a shield for that.

mckeemj

Sun, 31 Mar 2013 20:33:01 +0000

For larger motors it is typical to use external controller ( of your own design or commercial ) because of the complexity of designing reliable circuits to handle those power levels and for thermal reasons. These controllers are, typically, controlled via serial so they are easy to connect. Some sources would be pololu.com ( for smaller motor controllers ), roboshop.com( for up to 160A ), and trossenrobotics.com ( for other options up to 60A ).

Any chipKIT board can handle ten typical sensors ( be they SPI, I2C, analog or digital ).

Unfortunately, it's just not possible to construct a large complex robot ( of a new design ) without gaining some knowledge of electronics and programming ( and mechanics for that matter ) first... or along the way.

Martin Jay McKee

nightboy

Tue, 02 Apr 2013 20:42:24 +0000

By not relying on other people's work and designing your own circuits?

without gaining some knowledge of electronics

that is the main issue. :cry: high level programming(c,c++, c#,java) is not a problem but designing circuits is my problem.

So, the question is about compatibility. I found many motors controllers but most of them are not design for Chipkit boards. Even if I found one that could work with Chipkit boards, either it is expansive or very limited.

That's why I am asking others about their personal experience in robot building using Chipkit boards.

les1943

Tue, 02 Apr 2013 22:21:43 +0000

I would start on a small scale, get some ideas about the code required to drive motors, do you need over current sensing , limit switches, ramp up / down etc . you can breadboard small versions and then scale them up, plenty of info out there have a go.

Walk before run...