Little Black PI wagon

Let's Make Robots Robots
1

image_6.jpg2592×1936


image_7.jpg1488×1482

image_8.jpg1984×881




This is the beginnings of my little distributed computing bot. The brains will be a PI, the motor driver, a Dagu Minidriver. There will be several mini pros driving sensors like a gyro, compass, accelerometer and encoders as well as ir and sonar sensors. I will also have a ps2 teaching/remote control driven by another mini pro. All the uC’s and the PI will be linked together via an i2c bus and the PI to other computers with wifi. The main goal of this platform is to have a good easy to mod and room to work bot to test multi task based uC systems. This will be a multi deck platform.


The wheels on it are temporary and will have either track wheels or real tracks. I have not decided as of yet.


I will post more as progress is made.




UPDATE 2013-07-31


Added a few more componets to the base deck. Have a 3.3v supply, 5v and 5v for the RPI.


I think i have nailed down the power needs of the bot. To run the RPI and other controllers I will use two 5000mah lipos and a 2200mah for the two motors (will have to order). Going to order two current sensors and a voltage monitor and they will be placed on the lower deck as well. The wheels mounted now are what I plan to use for now.


Slow progress, but some progress.




*** UDATE 2013-08-04 ***


After thinking about it, I have moved to a pre-made chassis I already had that was just about the size I wanted. I have dropped the Dagu mini driver :( and replaced it with an uno and a rev 3 motor shield. The uno will be hooked to the Pi via an i2c bus, so same bat plan, just a new controller and hb.

Experimental platform

  • Actuators / output devices: 4 center shaft 100rpm gear motors
  • Control method: Rf link or autonomous
  • CPU: Atmel ATmega328, arm processor
  • Operating system: Debian GNU/Linux
  • Power source: 11.1v lipo x2 in parallel
  • Programming language: C++, C, Python, TCL
  • Sensors / input devices: encoders, compass, 3-Axis accelerometer, 3 IR proximity detectors, 2 SRF05
  • Target environment: home
This is a companion discussion topic for the original entry at https://community.robotshop.com/robots/show/little-black-pi-wagon
2

**Update??? **

Love to see where you are on this now!

 

3

I will post an update sometime soon.

I had some issues with a BIG i2c bus, but now have that under control. I was getting a lot of noise from the way I tried to implement at it first.  I have it set for autonomous via sonar and user control.  I have a compass and x/y/z module as well being read by one Mini Pro.  The Pi is not reading that yet (been working on another project)  The Sonar and R/C (PS2 controller) are both read by sperate mini pros.  All Mini’s are connected to the i2c buss as well as the Pi.  I need some 3d parts printed for sensor/modual mounts and will resume on this after I finally get my printer. 

4

Have had similar I2C issues. Motor noise (electrical) …

Have had similar I2C issues. I’m pretty sure it’s related to Motor noise (electrical) on mine.  Start moving, and lose all my sensors.

Put capacitors on the motors, and (knock on wood) things are working now.

 

So… are we building the same rover just on different platforms?  LMAO

https://www.robotshop.com/letsmakerobots/node/39791

 

 

5

We must be :slight_smile:

My main goal is a very distributed computing platform for a robot.  I want the PI as a master controller and many Pro Minis (small and cheap) as slaves.  I still have some details to work out like using mini’s as a pseudo interrupt controller. Have done some tests with good results, just need to write a demon on the Pi side to complete the test.

I had my i2c bus pulled up with a resistor and all the motors caped off.  My issue I think was the scew terminals I was using.  Vibrations and other factors where generating noise though them.  The only i2c item I have had issue with and the motors, was the compass.  I will put it on the upper deck on a pedestal a few inches off of that.

 I tried this in the early 90’s with limited succes using a RS-422 network with a Z-80 master and 8051 slaves.