Puma 560 simulation using pumakins

Fantastic!

I want to really thank you for that... I am not sure when I will get to start on a PUMA rebuild... but I know that eventually my controller will fail... so it is good to know that I don't have to start everything from scratch. I will contact the group and find the other individual with a PUMA... I'll let him know that when he van get about $850 saved for the Geckodrives... we should be able to set it up. (6 drives at $140 per drive.)

Thanks again!!

Hi Jerry,

what I forgot: My configuration is the MilliMMixPuma560.ini and not the other .ini files. It will start the control of the Puma via a normal parallel Breakout Board and at the same time the simulation window of the Puma simulation.
Btw: Two weeks ago my original Puma controller failed. There seems to be something wrong with the serial interface. I can no longer control the machine via external terminal.
I also thought about buying 6 Geko's for $850 but decided for UHU because I'm already using the UHU and the self-made UHU is only half of the price of the Geko (~350 EUR all together for 6 UHUs). The rest of the controller is using very cheap material (e.g. used 48V, 1KW server power supply for the high power servo supply). All parts of the controller together below 500,-- EUR.
When I find more time I will upload a short YouTube video.
There are still some obstacles - one is the internal correction moves of the hand joints JT4-6 what is not part of genserkins and the simulation.

Best regards, Thomas

I do have the 398H PUMA MANUALL... this also includes the schematics... it's a loose-leaf manual... so I could pull out pages and scan or photograph them if you do not have the schematics. There is a jumper between two pins... if missing, it goes from 9600 Baud to 300 baud... if that jumper came loose... that could Be a part of the issue.

Hand Joint - if the other 5 joints are working... we should be able to extrapolate to figure the DH for the remaining joint... in theory anyway. I really have to give credit to the original PUMA designers... they did a great job at the time.

I'll see if the manual has any DH parameters... I thought I saw something once... could be wrong. But if there...it will most likely be very useful.

Jerry

Hi Jerry,

I should have the schematics somewhere. I will first start to exchange the card with the one of a puma 260 controller in my basement (I keep for spare parts). But this has not the highest priority currently.
The issue with the JT5 correction moves looks a bit more complicated. I don't think that it will work with DH parameters. My first observation on the running original Puma (at this time it was still running) was: If the JT4 or JT6 moves the JT5 is from the TCP stable. But if you look inside the JT5 (or was it JT6?) is driven by the controller (even if you run in a joint mode only JT4 or JT6). This is interally calculated by the controller and it seems that the move is needed because of construction of the hand gearings.
I have to code the same somewhere in the LinuxCNC. At the first glance it looks like a proportional move depending on JT4/6. So if you have something with DH parameter I can test it if you can find it.

Thanks, Thomas

Hi jstoquica,

you IRB6 project sounds also very interesting. I guess the IRB6 is also running brushed DC servo's - correct? What voltage level? Does these servo's have standard encoders or resolver?
I just asking if I could use my universal controller also for an IRB6?

Many thanks, Thomas

Hi robottom,

In fact, I will replace all hardware components, because the investment time is smaller than test and repair some original components. The new components are 24VDC Servo Motors and digital encoders. 5 Gecko Drive 320X.

Can you send me more information about your controller.

Visit my blog, it|s more updating now.

plus.google.com/116960304375198732734/posts

Regards.

Hi jstoquica,

sure no problem. I reusing the complete PUMA 560 arm including Servos and Encoder. My controller consists of a normal DB25 parport board for 6 axis Step/Dir (as you are using) and 6xUHU servo controller: www.uhu-servo.de/servo_en/index.htm. The power supply is an old 48V server supply and some additional smaller power supplies (12V for the UHU servo boards and 19V for the PUMA breaks. That alls fits easily into a normal PC case.
The current controller is lacking of homing possibilities - e.g. not making use of the index-pulses and the potentionmeter of the encoders for absolute positioning/homing. Currently I have to home via scribe marks. These things will be added in a second phase of the project.
I'm planning to do a small YouTube video in the near future as soon as I have to first application working.
So my project is a minimum budget project compared to yours. Looks like you do it at the university - budget might be a topic but not that much as it is for my private one. This allows you to exchange more items. Exchanging the encoders sounds very natural (I guess the orginal IRB 6 is using resolver - correct) but why you exchange also the servo motors. Are the servos no longer working fine?
Did you ever consider to use something else than LinuxCNC - other open source operating software, e.g. ROS? Because I had some experience already with LinuxCNC I did not consider alternatives but I'm no longer 100% sure about this decision.

Best regards, Thomas

Hi robottom,

Are the servos no longer working fine?

Really, I don't know. The retrofitting methodology that I am working, doesn't test any hardware component.

Did you ever consider to use something else than LinuxCNC - other open source operating software, e.g. ROS?

I am working with open architecture (OA) concept, then I am mainly using LinuxCNC because is GNU and add value for the OA concept.

Regards.

Hi guys,

today I could find time to do a little video of the project and uploaded it to YouTube:
There is a little intro to the controller (prototype) and a first test.

Thomas

I am very excited to see this. My controller is working, but starting to show signs of failure.

If I can follow your work and convert this to work properly with my machine... then it will be possible to create new software to use the STEP-DIRECTION outputs to control it as well. Great job!!!

Please follow-up with the interface MESA or PARALLEL PORTS that you are using as well as the servo boards you are using. I will want to gather as much information as possible, then eventually convert my machine to the new controller as well. My machine is a bit older... I am wondering if it has the same encoders or not. I hope so... that is a big part of making it work properly.

Jerry