Did you select correct tool?
Nonsensical TCP behavior
-
Wall-E -
February 7, 2022 at 2:29 PM -
Thread is Unresolved
-
-
Did you select correct tool?
Yes, I did.
I just received a linear guide rod which has a very small runout, and made a new calibration tool with it.
So it's back at monkeying the robot around to define a tool and hopefully get something that makes sense.
I'm trying to reach the local KUKA people to borrow an EMD too.
-
Is your tool pinned?
-
Is your tool pinned?
Ummm... what does that mean ?
-
Mounted not only with screws, nuts and bolts. There should be at least two precise pins to fix the position precisely.
-
as hermann just explained
Your tool looks very heavy and very long
and one addition to that:
for every adapter
- flange a6 to tool holder
- tool holder to tool
you should use them
also have you checked the required torque of the screws?
-
screws holds the tool
pin holes and pins hold the tool in place
The calculation of the tool data should be done afterwards
-
Mounted not only with screws, nuts and bolts. There should be at least two precise pins to fix the position precisely.
Nope, there are two remaining holes on the baseplate, and underneath, threaded holes on the plange.
-
also have you checked the required torque of the screws?
Nope.
I'll look into this, but the guys at Eurobots know what they are doing ; I "suppose" they have tightened the nuts correctly...
-
you have not given us the tool data and did not pin the adapter plates
Your result for this combination is just perfect - be happy with it
The pin holes are not made for fun, they made to get a better result
-
The pin holes are not made for fun, they made to get a better result
OK Mom, I'm sorry, I won't do it again.
-
do not apologize, we want to help you and therefore come up with all problems which may happen and this could be one of them
-
OK, sometimes my perfect English is sometimes very limited
-
I'll look into this, but the guys at Eurobots know what they are doing ; I "suppose" they have tightened the nuts correctly...
Nope. Torque doesn't matter -- any clearance hole for a bolt still leaves some small amount of "slop". Over a few hundred mm, that tiny amount of slop can add up to several mm of slop in the TCP. And a dense, heavy load like a spindle can easily move on those bolt holes every time you rotate the wrist, even if it feels 100% solid when you grab&shake it.
The rule is, bolts for retention, pins for positional tolerance.
You should also check every link between the A6 wrist flange and the TCP -- there should be bolts and pins at every junction. Robot<>bracket<>spindle.
-
the nonsensical part is still not clarified - what is the selected coordinate system, what is the tool value, if the tool direction is set correctly, if the axes were correctly identified (jog X, Y and X to confirm) and really match the expectation.
with those things still in the wild, i would say checking if tool is pinned is a small fish - even if all EOAT bolts are lose and tool is sloppy, it will not make the tool turn around 90deg or so. if the tool is firmly mounted (even if not pinned correctly) then measuring of TCP will not be a problem (unless tool is removed and reattached - that would require new measurement).
-
the nonsensical part is still not clarified - what is the selected coordinate system, what is the tool value, if the tool direction is set correctly, if the axes were correctly identified (jog X, Y and X to confirm) and really match the expectation.
OK Panic, although I'm slightly moronic, I think I know what I'm doing here, and I have described all the things that you cite.
As well as you are prompt to point to documentation, I will kindly point you to my very first post where I have taken great care to explain what I was doing, which coordinate system, etc... with bonus photographs too.
-
I mean no offense, the way i look at things, someone asking for help is not doing something right and not sharing details that could make this easier to resolve.
Whatever the situation, fresh look at it is supposed to help. The thing is that fresh look means there need to be something to see.
And I still do not see (for myself) what the tool value is, what the tool number that is, what tool is selected, what exact coordinate system is selected, what the tool length is, if the tool is mounted in correct orientation, if tool_data is matching or way off etc.
This means no way to offer help or determine if the values even make sense.
What i see is someone with no formal training asking how to get ABC values in for the tool and things like that are dead giveaway that user is struggling with much more than shown. Because i am trying to help you, it would be nice to verify few things... For example screenshot showing value of tool_data [n], screenshot of selected tool and coordinate system, photo of the tool when robot is cannon position, have the photo show measure tape next to tool for scale reference...
What i am trying to say is that you do not see the problem description my way, and you are assuming that your posts are detailed. Trust me, they are not, at least for me. And i have spent years teaching, using robot, programming providing support, troubleshooting etc. I am used to helping people with little or no experience and i know what i am asking and why. If the other side knew that too, they would not be the ones needing help, they should be able to resolve things on their own.
Just trying to help. But If you think that my posts are not helpful and you dont want me to participate in this, that is totally fine by me. You are more than welcome to continue doing things your way.
-
OK... I'll just start from scratch and document everything.
I don't want to get in bad terms with a fellow Canadian.
-
Yes, we all know how nasty Canadians are. I mean, their national sport is hockey!
-
Careful... Justin was in my class when I was 12 in Ottawa.
Just a phone call and he can send a trained beaver to bite your rear !
-