Inconsistent tool speed with close points

nsadding

I am using a IRC5 6400 for machining complex surfaces and am having difficulty maintaining consistent TCP speeds. I am using a third party software to slice/ auto generate G code from STL file and then using RS CAM converter to generate Rapid code.

The Slicer software produces a lot of targets, very closely spaced at fractions of a mm apart.

The problem is that the robot tool speed varies wildly when executing the machining program.

I suspect that the robot is struggling to execute the Rapid instructions fast enough and is hesitating whilst it is processing a buffer of code with very short movements.

Are you aware of any fix or workaround that may assist to minimize the speed variations?

Typical machining speeds are about 200mm/s and robtarget spacing at between 0.2 to 1 mm.

Any help would be greatly appreciated.

graemepaulin

Yes this will be giving the CPU a work out.
Check that your zones are not larger than half the distance between the robtargets - if larger than this the system will consume even more CPU power to reduce the zones on the fly.
If the CPU becomes overloaded you should be getting warning/errors.

xerim

Also might be worth creating your own speeddata parameter to control the speed at which the TCP moves and the reorientation speed of the tool 

lemster68

There are system parameters that you could change to read ahead more, but you have to be very cautious in using them.  One of which is prefetch time, in motion parameters.  You could look into high interpolation priority, and path resolution.  Finally, Queue time.  Study them all carefully in the system parameter manual.

lemster68

You could also use the rapid instruction AccSet to reduce the acceleration and deceleration of the robot.

nsadding

Thanks to everyone for very valuable feedback.

I have a few trails to follow now.

I'll be trying out each of your suggestions over the next couple of days and will let you know what I find.

Thanks again for your very valuable and timely suggestions.

nsadding

Hi All.

Just a quick update to close off this thread.

I have done a lot of work and trialing to minimize the robot speed variation.

The take out is that reducing zone size to less than half of the robtarget spacing and applying a reduced pathresol value (in my case 50% of default value) has improved the speed regulation to an acceptable level for this machining application. I am sure that there is a lot more to learn about optimizing the robot speed and I will continue to study this problem. For now, I am happy that we have a workable program.

Thanks again to all of you for your very valuable inputs. I truly appreciate your help.

lemster68

Great!  Thanks for the update.