Comarc Issues

  • I was hoping that i could ask a couple questions in regards to the comarc.


    My company just purchased an aw2000 cell with touch sense and comarc. We did the install ourselves and everything seems to be working somewhat. There are issues we are trying to troubleshoot.


    Im interested in the comarc hardware installation actually. I know things change from version to version so im seeking some clarification.


    We have a Comarc IV. Its mounted to the side of the dx200 controller. I am assuming that this serves a dual function as the starting point detection unit and the tracking.


    The issue we are having is with the comarc. It doesn't seem to track very well. I've had it come right off the seam a few times.

    We've had a motoman guy come up and set it up for us and give us the run down so we are pretty sure we are setting up the settings right. u/d matches welder current output L/R =0


    We have an abb with seam track aswell and i remembered after we got it we had some tracking issues and it turned out to be interference from the + cable to the current sensor cable.

    I decided today to take a look at the way our maintenance guys installed the new robots tracking sensors and it left me with some questions.


    I was looking through the comarc manual at the wiring setup and while some of it looks right, other things leave me wondering.


    Im looking in manual 165557-1cd


    First off.. i don't see a current detection sensor anywhere....built into comarc iv unit?

    Secondly, the configuration diagram of the starting point detection unit in the manual matches our setup except for se1 and se2. These connectors are on the back of the unit but there are no wires connected...Is this correct?


    The warnings on pg 2 of the manual, down at the bottom there is mention of interference from the welding cables and that we should separate these cables from the control circuit system. DUH

    The only wire coming out of the comarc iv is E. This wire runs to the welders neg terminal. Our maintenance guys zip tied that with the + and the - coming out of the comarc unit. Would this be a control circuit cable that needs to be isolated from the positive and ground.


    After reading this description of our system, are there any other cables that should be isolated that im unaware of.

    The touch sense works fine and the seam track seems to function just not well. I get better welds if i remove the comarc and just weld it normally.

    We are talking about straight forward 3/8 fillet welds on a gentle 5-10 degree slope. 2-5 degree push on 1/2 inch semi-clean steel. They go in great without the comarc but with is a different story.

    I'm even using the same weave files without the comarc as i would with.


    Its smells of interference but im not sure.

    Any clarification would be greatly appreciated.

  • AD
  • For COMARC you need to be at the leading edge of the puddle. I would suggest a 15 degree push angle to guarantee that this is the case.


    What kind of wire are you using - solid wire, flux-cored wire, metal-cored wire?

  • I fit up 2 parts before the break. I welded one without comarc. It welded perfectly. Then I copied the program and edited it to use comarc. Kept all the same settings, arc files, weaves. Justed edited in comarc instead of weaves alone. The second part is junk.

    When I welded the first part I took the current output of my welds and edited into the comarc u/d for the second. The l/r was 0 for all welds. The comarc is causing the torch to climb the wall. In some cases over 1/2 an inch. In others less but still enough to cause cold lap in the toe. Every weld as well. Not just a few.

    We have motoman returning on the 6th with a new comarc. I’m not sure if it’s the comarc or interference but there is something definitely wrong and my money is on the hardware. We were having issues before with it doing some strange things. In some instances during dry walk through, the robot would jump and trigger a stop. Very odd behaviour. Motoman came and looked at it. Said they had never seen this kind of behaviour before. They ended up dumbing down the sensitivity of the comarc and changing up the my weave frequencies. Some of my frequencies were too high for the weaves I had setup but I wasn’t using those weaves for any of the welds that were suffering from this jump issues but it was something motoman seen and criticize.

    It stopped the jumping but I’m starting to think the two issues might have been related. We shall see. I know it’s easy to say it’s a process issue ( put it on the customer) but when you see behaviour that just isn’t normal and your tech guys haven’t seen it before and you have to send backups to Japan to get analyzed, maybe you should consider its more then just a setting and it’s possible it could be hardware. It blows my mind that they would think that dumbing down the sensitivity to correct an issue they’ve never seen before was the way to correct the issue. It doesn’t explain the root cause and that’s what they needed to address. Not just find a settings fix that masked the original issue.

  • Sounds like you need to start from scratch with the setup. 0 for L/R and true current for U/D are not the correct settings.

    Collect PC data and set the parameter defined.

    Collect sensing data and register it to the COMARCON instruction.

    ensure the travel speed is correct for the Hz and AMP in the weave.

    Use the same grounding configuration for the setup and testing.

  • I'll be honest with you; probably 50% at the very least - of anomalies I run into are COMARC-related. It doesn't matter if your arc files are perfect, it doesn't matter if it's run it properly 50 times, sometimes when it's doing it's game of catch the current it just seems to go "NOPE" and go freestyle on me.

    That's not to say it fails half the time, it definitely works when it works which is more often than not, but sometimes it doesn't. Though it does sound like you've got a fault in the setup for sure, it should lay relatively flawless weaves at least occasionally for you.

    I used it a lot when I wasn't as familiar with the robot and its touch sensing capabilities - now that I am though, I prefer to use those in abundance because otherwise using COMARC we have to babysit the thing so it doesn't bork the whole part.

    Weirdest bug I've found is that if we're doing a coordinated weld path, when it rotates more than a relative 180° the wave will just, roll like 90° and weave up instead of out. Baffling stuff.

  • This is not a bug in the system. When you are using coordinated weld paths you need to teach SREFP at every 90 deg. As you roll the torch the ref point 2 changes so the weave angle also changes.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account
Sign up for a new account in our community. It's easy!
Register a new account
Sign in
Already have an account? Sign in here.
Sign in Now