14 years old, RJ3iB controller, has about 4200 running hours on it. I don't have the small hole accuracy option. Our product is usually 5-8mm thick, but I've cut 12mm with no problem. I can hold about a 1mm tolerance. My usual hole sizes range from 1"-2.75."
Small Hole Accuracy
-
popbot -
April 4, 2017 at 1:30 PM -
Thread is marked as Resolved.
-
-
Is it possible to post a photo of a typical hole. Something I can take back to my group. I'm in a tight spot here a photo would go a long way helping me with my group.
-
Here you go. 1 3/4" hole viewed from both sides of the product. It looks a little bit misshapen, but that's because the surface it's drilled into isn't flat - not a problem with the hole.
-
looks like the material from the first photo in this topic gets deformed in the process - I would say that is the cause of these irregular shapes of the hole, especially of these "horns" inside. Of course, the faster you cut, the bigger the deformations and the worse the quality.
OP, while cutting, does the robot move in the same or opposite direction as the tool? Have you tried reversing the direction of the circular motion?
-
Yes the tool spins clockwise as well as the cut direction. For the record, I have tried to cut counterclockwise and got the same result.
-
Popbot: Have you tried a fluted "end mill" type cutter? Looks like you're using a carbide burr. The fluted cutter should help reduce the force needed to make the cut.
-
The rasp bit I posted has two flutes on it, just cant be seen in the photo. This rasp style was recommended by two different tool manufactures. I am defiantly interested in how the bit TitusLepic is using would work. (Posted below)
-
The TitusLepic bit is certainly worth a shot. Try it out and let us know how it works.
-
Anyone heard of the Dean Magee method of mastering? Is it more than just lining up the witness marks? How so?
-
Hi,
Never heard that name before, There is a Fanuc document titled Accuracy Improvement
I can´t find it in this laptop
Basically there are multiple solutions to reach the same position
You record one position, then you modify the configuration and compare both
Ideally both solutions should move to the same position
if not, is because a small calibration errors (even brand new robots are not perfectly mastered)
The document provides instructions to calibrate J2,J3,J4,J5 -
I programmed a macro too cut circles, you only need to teach the center point normal to the surface
Then you specify hole diameter, cutter diameter, clearance distance, and cut depth distance
The macro calculates all movements to a point above the piece, then move down to the specified cut depth into the material
Then cuts the circle with cutter diameter compensation, finally moves clear of the material
Since the macro calculates all points, is necessary to good masterization and good TCP -
robotero, That sounds like the procedure we now have. Our robots are apparently too old for the on-board tools like Robocall and whatnot. This Mcgee method looks like its probably better anyway. I also have a macro I am working on, one that can make a hole out of round on purpose in hopes that the result would be round.
-
I am having great success doing some servo tuning. I'm getting round holes!
-
I am having great success doing some servo tuning. I'm getting round holes!Sounds interesting... could you explain how you are tuning the servos?
-
Consultants have had us change some values in servo parameters. I’m not sure exactly what I’m adjusting, servo gain or something. Its apparently proprietary so I’m not getting a lot of explanation. It has had a definite effect on how the servos handle the backlash.
-
Good to hear you're getting round circles.
-