IRC5 M2004, how to install a 7th axis?

    Hi,


    I have a IRB6400RF IRC5 M2004 standard robot with nothing external connected.


    Is it reasonable for an amateur to buy used parts and install an external rotary axis to this system? Where do I start?


    Any recommended hardware?

    Do I need to add another drive in the controller?

    Do I need extra options in the controller software?


    ABB says it's not possible because my system is "obsolete". "Better to buy new". But the system duty time is 1800 hours and everything is working fine, so that seems stupid.


    Care to give me some advice?


    Best,

    Joakim

    Quote from Skooter

    Depends on what you want the rotary axis to do.

    I'm thinking a flexible setup. I have a model workshop built around the robot with a turntable built from a skylift bearing. I also have a DIY "lathe" kind of rig based on an old indexing rotary table. Both of these are currently manually positioned, but can easily be driven by a motor.


    My idea is to have one motor that I can move inbetween the two rigs because I will not use both at the same time. I am milling styrofoam, polyurethane and wood, soI suppose I need a somewhat powerful motor.


    I have seen ready made ABB motor and cable packages, but none for sale and no information on how to install it.

    Just using the robot as a motor controller will be expensive as you will need the options enabled in the software plus the additional robot components are expensive on the open market.

    I'm cheap so I would consider repurposing a garage door opener.

    Quote from Skooter

    ...as you will need the options enabled in the software...


    Thanks Skooter for confirming this. That alone rules out integrating another axis and the other factors are decisive as well - the costs are just overwhelming.


    Interesting idea about the door opener. I will probably sort it out with a stepper motor, a PLC and a digital IO card. I have a few IO cards already, so it's the cheapest option.


    A coordinated axis would be nice but in practice it's probably more fancy than crucial.

    As stated, the issue is the drive units that are in your robot - "Drive system 04".

    These are obsolete, so getting an additional drive unit is going to be expensive to buy, and will have to come from a 3rd party, not ABB. So the cost of the drive unit, and then adding a keystring to enable the external axis functions, and the motor / cables etc is not cost effective.


    VFD with a motor and some analog inputs/outputs would work nicely for what you are looking to do.


    Coordinated axis are cool, but mostly not needed in most applications.


    Have fun!

    Quote from D_Oster

    VFD with a motor and some analog inputs/outputs would work nicely for what you are looking to do.

    Thank you for your reply! I actually have some spare parts from this robot when it had something like that attached. The table hardware is gone and the motors too, but the seller didn't know what to do with the control cabinet so he gave it to me. Here are some pictures:


    cabinet1.jpg


    sew.jpg


    cable1a.jpg


    cable2.jpg


    digital_io.jpg


    analog_io.jpg


    The robot has a position switch and a collar that activates it where you want to. My guess is the robot had a grinder or something controlled by the green VFD (labeled "milling motor") and was programmed to swing away from the table to activate the switch around A1. The table turned (controlled via the SEW drive, it has a "Axis 7" label) and the robot went back to work.


    Any suggestions on how I can use these parts for a new turntable? Would it be controlled similarly, with switches? How would the precision compare to a PLC driving a stepper motor?


    What does the SEW drive do anyway?


    Sorry for my amateur questions but I'm eager to learn.


    Thanks,

    Joakim

    The SEW seems small at 1.5KW. I guess it depend on how you gear it.

    Find manual at https://download.sew-eurodrive.com/download/pdf/10538313.pdf

    The position switch setup is handy to ensure things move or don't move when the robot is rotated to a certain area.

    The digital & analog cards are there for whatever you need them for.

    If the robot still has an old backup of the previously setup, that could give you a lots of information on how it was setup.

    Quote from joakimmalare

    Thanks Skooter for confirming this. That alone rules out integrating another axis and the other factors are decisive as well - the costs are just overwhelming.

    What were you quoted for enabling the axis in the software?
    I'm looking at buying a used IRB 7600 with the IRC5 M2004 controller and would like to use a rotary axis. Like you I already have an existing rotary table that I use with another CNC machine (custom made for turning 20ton blocks of stone).
    Thanks

    Quote from greg9504

    What were you quoted for enabling the axis in the software?

    The software part wasn't specified in the quote I got. I was given an estimate for installing an A6 motor and drive as a 7th axis, motor included. Around EUR 7000 at the time. It was a good price considering the cost of hardware but I'm not 100% they had accounted for the cost of software upgrade.


    When I asked ABB directly they recommended a controller upgrade (with added axis) for something like EUR 30000. I believe that didn't include the actual turntable.


    I'm going for a DIY solution; a PLC and a stepper motor. Total cost less than EUR 2000.

    Thanks for the reply.


    Quote from D_Oster

    Coordinated axis are cool, but mostly not needed in most applications.

    Is this really true for what you want to do? It sounds like you want to mill with the robot?

    With your solution will you only be able to index the part (perform machining on one surface, stop machining, rotate part, start machining on next surface)? Or will you be able to do coordinated moves for milling a 3D shape? That's where I'm confused. Without the robot controller knowing of the turntable axis, I don't see how coordinated movement can work.


    If you are bored, here's how I used a slew bearing for my setup:

    External Content youtu.be
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.

    Plain 24VDC brushed motor, very old Advanced Motion Controls B12A6D Brushless PWM Servo Amplifiers ($50USD) and a 48VDC power supply. The CNC controller uses +/-10v to the drive.

    Quote from greg9504

    With your solution will you only be able to index the part (perform machining on one surface, stop machining, rotate part, start machining on next surface)? Or will you be able to do coordinated moves for milling a 3D shape? That's where I'm confused. Without the robot controller knowing of the turntable axis, I don't see how coordinated movement can work.

    Yeah I do all sorts of milling and an indexed positioner would work fine. Coordinated movement isn't really necessary since the operations can be split in a few runs around the model. The most common use of coordinated movement is rotary finishing with infinite rotation and yes, that is convenient - but it doesn't justify the cost.


    Currently I have a manual turntable built on a skylift bearing that has a geared ring, but I don't have a drive on it. I split the milling and turn the table manually between operations. To make it automatic all I have to do is to attach a gear and motor with a drive and a PLC ( https://velocio.net/ace/ ) connected to the robot controller via digital IO. Four signals from the controller gives me 16 positions. The robot moves away, the PLC does the turning and gives a signal back when it's done.


    In my CAM software I have a workpiece holder oriented according to the rotary axis and each rotary position is essentially a new workpiece setup (rotated around the holder Z axis). Before the next operation I add an auxiliary command that outputs the relevant signal combination. So it's a dumb setup, yes. The controller doesn't know about the rotary axis, it's just switching workframes. The CAM software doesn't know about the external axis either since it's set up as manual repositioning of the workpiece.


    To make it a bit less prone to failure I suppose it would be a better idea to add an external axis in CAM and derive the axis position data from there in the post processor. I haven't figured out how so until then I can go with the extra command.

    Quote from greg9504

    If you are bored, here's how I used a slew bearing for my setup...

    That's very nicely done! Way beyond my skills. I will look more closely when I'm not at work, but it's cool to see DIY stone masonry CNC builds. I worked at a stone masonry as an CNC operator for eight years and did all the programming there for 3, 4 and 5 axis machines. Profile milling, polishing, 5-axis sawing and some lathe saw / milling of sculptures.


    Great work!

    Quote from joakimmalare

    To make it automatic all I have to do is to attach a gear and motor with a drive and a PLC ( https://velocio.net/ace/ ) connected to the robot controller via digital IO. Four signals from the controller gives me 16 positions. The robot moves away, the PLC does the turning and gives a signal back when it's done.

    This is interesting! I'm curious, what happens at the edges where you split the model? When you split the model do you overlap slightly? So there is no obvious transition in the machined surface? If I understand correctly, instead of using one model in the CAM, I'd slice up the model into N number of sections, run each of those through the CAM, then run them as N distinct jobs on the robot? Is splitting the model even necessary or can the CAM handle that?

    This solution seems like it could work well for me. Thanks for sharing.

    Quote from joakimmalare

    NC operator for eight years and did all the programming there for 3, 4 and 5 axis machines. Profile milling, polishing, 5-axis sawing and some lathe saw / milling of sculptures.


    Odd that we both posted here! What I've done isn't too complex, I'm sure you could have done it to, especially with your background. Once I started looking at 5 axis machines for milling stone they all had the same problem, not that large of a working area. And the ones that had any size were super expensive. That's why I started to look at robots. They seem a perfect fit for large stone work.

    Again thanks for sharing this solution. I going to try to find more information on indexing...

    Quote from greg9504

    This is interesting! I'm curious, what happens at the edges where you split the model? When you split the model do you overlap slightly? So there is no obvious transition in the machined surface? If I understand correctly, instead of using one model in the CAM, I'd slice up the model into N number of sections, run each of those through the CAM, then run them as N distinct jobs on the robot? Is splitting the model even necessary or can the CAM handle that?

    I'll post a screenshot next week what my workflow looks like. Once figured out, it's easy to repeat.

    So, here are some screenshots. The first explains the setup with the software workpiece holder centered at the hardware bearing.


    screen01.png


    screen02.png


    Each of the setup stages "01..02..03" has a rotation value for the workpiece holder (like a lathe chuck).


    screen03.png


    Above you can see "Base setup" says "GlobalCS +(C-180)", so the workpiece is turned 180 degrees.


    screen04.png


    This is a polygon model with the mesh split in advance in a suitable manner for milling, so the operations only work in designated areas.


    screen05.png


    Some parts were too tedious to split nicely, so I have a method to shrinkwrap curves on the scan data. These are split at their intersections so that you can combine a few to make up a field for milling in different angles.


    screen06.png


    Finally, in the code from a different project (similar setup) you can see how the different setup stages result in their own respective work frames, defined initially at the top and referenced in each operation as necessary. Here it becomes obvious where you would want the custom code inserted for the automatic turntable rotation.


    I haven't made any example operations for the turntable commands for you to see but I suppose it wouldn't make sense unless we use the same CAM software.

Advertising from our partners