Adapted OWI 535 Robot Arm as Picaxe Robot with psuedo CNC controller

Anobium

Senior Member
Completed robot project. I will post information over the next weeks and links to my microblog.

The robot is intended to educate give hands-on access to robots operating in an CNC environment (in terms of the G Code language).



The robot is based on the OWI Robot Arm with adapted motor controllers. The PC is the psuedo CNC controller with a single 18M2 microcontroller controlling each axis. The manual controller has been converted from analog to digital and has additional buttons to provide control modal controller and an emergency 'stop' control for the motors and the picaxes.

I have documented the process of adaption and will post photos and videos asap.

http://www.youtube.com/watch?v=5wiKAZYmUq0

The have create a website to post all the details of the projects and all my related resources, see here

Thanks to everyone who helped me to complete this project. As I document I will thank everyone again.

Anobium
 
Last edited:

Anobium

Senior Member
Solution Overview Documentation

High Level Architecture Documentation Released. 31st July 2011


See here for the high level architectural documentation for the modified robot. The document shows all the components and examines the function of each component.​



Low Architecture Documentation In Production @ 1st August 2011
 
Last edited:

Anobium

Senior Member
Gearbox modifications and other robot mods

{Draft 0.01}
The gearbox box modificationS are listed below with links to sources and photographs. The modifications are supported by the Picaxe 18M2 in terms of counting gearbox movements and rotational position of the gearbox.


The modifications goals are to enable movement feedback and positional accurancy.

1. The first modification was an slotted opto sensor. See http://www.instructables.com/id/Modifications-to-Robot-Arm-for-Opto-Coupler-Feedba/, this details the methodology. See later comments on the findings and issues with this approach. The Picaxe will be used to count gearbox movement.
2. The additional modification is a position sensor and an 'location indicator'. A QRD1114 provides rotational position. As the gearbox has limited rotation, in terms of degrees of movement. The Picaxe will be use to locate a locate the 'location indicator'.

The photograph below shows the completed gearbox modicatio, the gearbox was lubricated and closed as per the OWI build process.


The complete build process will be posted {here}.​
Additional Robot Modifications]

The complete build process will be posted {here}.
 
Last edited:

erco

Senior Member
Brilliant! What a great job fitting those encoders inside the case. I'm even more impressed now. :)
 

Anobium

Senior Member
I have created a website to support the project. See here for the website dedicated to this project.
 
Last edited:

Anobium

Senior Member
Nice work!
Can't wait to build my own 6 axis Dc servo Robot + Picaxe cnc controller.
Is possible to use psuedo CNC controller for kuka robot?
Thanks for the feedback.

If we can submit the pseudo CNC software for full safety checking, then an end to end engineering review and a little more testing we may have a cheaper option for manufacturing! Joking apart, the intent is educational - you can learn vectors, learn co-ordinate systems and have some fun whilst using the robot…
 
Last edited:

erco

Senior Member
Anobium: You are using one 18M2 per axis, correct? Is that overkill? Seems like one 18M2 could handle multiple axes & encoders. I'm just starting with the 18M2 & 20M2, hoping that the multitasking is fast enough to track wheel encoders and navigate on a basic diff-drive bot. I've had excellent results with a non-multitasking BASIC Stamp 2, and I'm hoping that a 20M2 will do much better and allow much more flexibility. Here's my BS2 robot: http://www.youtube.com/watch?v=PX0IhUqnwrk
 

Anobium

Senior Member
Anobium: You are using one 18M2 per axis, correct? Is that overkill? Seems like one 18M2 could handle multiple axes & encoders. I'm just starting with the 18M2 & 20M2, hoping that the multitasking is fast enough to track wheel encoders and navigate on a basic diff-drive bot. I've had excellent results with a non-multitasking BASIC Stamp 2, and I'm hoping that a 20M2 will do much better and allow much more flexibility. Here's my BS2 robot: http://www.youtube.com/watch?v=PX0IhUqnwrk
Nice robot!

The 18m2's are coping multiple interrupts and all the ports are usehero control a single axis. The high level architecture of this type of multi axis robot is different from a wheel/tracked robot. The requirements of the Picaxe microcontroller were:
  1. Geabox Sensor #1 - Movement sensor
  2. Gearbox #2 - Home Position
  3. Comms in and comms out
  4. Motor control - PWM for feed control
  5. Motor control - Direction of motor
  6. Interlock control - Interlock with all the controllers
  7. Stall detection - Detects motor stall conditions
  8. Axis ID - (3 inputs to cope with up to 7 axis of movement - tested 5 axis of movement
The port usage is shown in the attached PDF.

An essential part of the solution is that all the motor controls are the same. The same Picaxe code, same components etc. All configuration is through the Axis ID ports or via the PC software. The choice of Picaxe chip was 1) predetermined by the input/output requirements, 2) my design goal to ensure the motor controls were the same and 3) provide reuseability for anyone wishing to build a similar multi axis robot project.

I hope this helps.

Anobium
 

Attachments

Last edited:

erco

Senior Member

Anobium

Senior Member
@erco Thanks for the feedback.

I think I will post to the hackday web site. I just need to find some time. I been contacted by a few people and the project has given other inspiration - I am pleased.

:)
 
Top