View Full Version : Corner problem

06-18-2006, 01:31 AM
Can some one explain why am I getting this corners?
See pic below

I have a PRT96 with SB3.4.25 control software.
Checked all the bearing no play found


3805 (26.5 k)

3806 (3.7 k)

06-18-2006, 02:03 AM
From Ravi's .dwg file

Ravi, that problem has nothing to do with electrics, electronics, controller or software. It is entirely due to mechanical issues. The usual suspects are:
- V-rollers on Z-slide need tightening (the slide may have worn hollow areas)
- pinion gears loose on motor shafts (grubscrews)
- router mounting bracket is loose
- the v-rollers of the y-car are not in positive contact with the rails
- and a whole lot more....

To find the culprit, turn the router/spindle off and grab it firmly by the collet. Push/pull hard (80 lbs) in x and y directions and see what moves. (if the z-slide is worn, you need to check that at the right z-depth).

06-18-2006, 03:05 AM
Check also that the vertical inside lines do stop where they are supposed to.
If you offset the lines the bit radius, and didn't trim off the excess, you could get a similar problem.


06-18-2006, 08:47 AM
You are going to have to explain yourself better. What you show in your picture, (it is a picture of your results right?), appears to be a mechanical issue of your machine when it transitions from moving negative-x direction to positive-y direction. That's the assumption based on what little you have presented us with.

Other things that have caused problems similar:
1. set your toolpath resolution to a lower number before making toolpaths in parts wizard
2. you made your toolpaths yourself by offsetting the shape, and the original shape doesn't have a truly square inside corner, elipses, curves, and radius corners can be dodgy when offset or outlined.

06-18-2006, 10:47 AM
Does the problem only show up in the bottom left corner? Try cutting a rectangle and a circle and post the results.

06-18-2006, 01:33 PM
If the problem is seen only where shown and not here...


...then only on axis is concerned. I'd bet on a loose pinion... or a worned out gear box?

06-18-2006, 05:27 PM
Thanks guys
I am going to check one by one and let you know the results

thanks again

06-22-2006, 05:23 PM
We've got the exact same problem.

Please post if you find a solution. Ours is brand new. I don't think we have any worn parts and it just popped up.

06-22-2006, 08:23 PM
Just noticed I have the same problem. I never noticed this before. I have the latest software and a new Milwaukee router. All my drawing is done through Artcam Insignia and KCDW cabinet design software. Seems like this has to be a glitch in the ShopBot software because it's showing up in different spots on the panel. I don't think it would be bit deflection because I'm using the same bits I've always used and the same toolpath strategy. Maybe one of you smart guys can figure this one out. Kip

06-23-2006, 07:23 AM
I had a similar problem when i first got the machine.it turned out to be flex in the machine. try cutting the part in the opposite direction to see if the notches mov to another spot.

06-23-2006, 06:17 PM
My problem is occuring in different places on the panel. As I stated in my previous post, it does not appear to be deflection in the bit or the gantry because of where it shows up on the panel. I've cut out hundreds of panels and I've just recently noticed this problem for the first time. Thanks, Kip

06-23-2006, 07:07 PM
If you're 99.999% sure that's it's not related to the CAD/CAM process, then, like I told Ravi this week, your best friend to troubleshoot this kind of problem is a dial indicator. Get familiar first on how to use it then you'll be able to point out ALMOST any play that your tool might have.

06-23-2006, 08:01 PM
I had a similar problem a year or two back, it was related to slop in the rack and pinion setup. In my case it was the Z axis that had a little play in it. Adjusting the R&P on that axis solved the problem.

06-24-2006, 12:50 AM
Tried everything I still have the problem.
Only thing left is to move the router closer to the Y motor.
Hope this might solve the problem

06-24-2006, 01:13 PM
If this is the problem, you should see it on the dial... I don't have this problem (as may others) so you shouldn't have it either.

06-25-2006, 01:38 AM
I noticed this problem yesterday.
I was cutting out the green part, the red part is waste.
The corners of the green part are perfect right angles. This means that after making the 'notch', the tool backs off to where it should be.
The tool over-runs, then backs off. It happened at all 4 corners. The arrows show direction of travel. When you think of which way the force is acting against the tool, this seems to preclude any flexing, and suggests, maybe, that the software' or drivers' is 'over running', then correcting themselves. I feel if there was flexing, then the corners of the green piece would be shaved off, not the outer red piece.

(The size of the notch is exaggerated in this drawing.)



06-25-2006, 02:03 AM
Theoretically, if it is a case of the system "overshooting" before turning a corner, it can be that the speed has not ramped down low enough before the direction change.

06-25-2006, 10:52 AM
The only occasions I've noticed this kind of problem was the Y pinion loose and on a VERY aggressive cut in plastics and aluminum where I made toolpath error and the tool machined a full kerf quite deep into the material and/or feed and/or RPM was way off. In those late case, I believe the rake angle of the cutting edge was actually driving the tool bit further into the material from some sort of pulling action. It was more obvious at the corners but the whole part was too small (conventional cutting). This late as been rather rare and is partially relate to the flexibility of the Bot; if one machine as some part loose then it'll get worse...

You may want to triple check the torerance of the toolpath if any there is in your CAD/CAM package.

06-25-2006, 12:18 PM
See Paco it's not just me,
I agree with Mike that this has to do with the driver, As I mentioned earlier I checked with dial gauge and there is no play in any axis.
There is a bit of flexing but that will not cause this kind of an error

See picture below all four corners has the over run, Probably about 0.025" over run

Can some one cut a piece like this with 1/4" bit and check the corners


06-26-2006, 12:40 AM
How fast are you cutting with this 1/4" bit? What are your ramp speeds set at? What direction are you traveling (conventional or climb)? If you're slamming into the corners (especially for a y axis move), you can get marks like that from the abrupt stop and go causing the machine to shake. If you are climb cutting with a 1/4" bit with a 1/4" shank, you would be getting the marks in the corners like Paco's drawing shows due to the bit deflecting. Try air cutting the file and place your hands on the carriage, can you actually feel the overtravel?

If you're sure it's not mechanical in any way or that the code in your .sbp file doesn't give any evidence to any moves beyond what is supossed to be there, then check your speeds/ramps. Try the same cut with a 3/8" or larger bit. Also, try cutting a rectangle with the "Cut Rectangle (CR)" command or hand code a rectangle cutting file to see if you get the same results.

06-26-2006, 01:07 AM
Ravi, I can't figure out your picture - what am I supposed to see there?

06-26-2006, 03:25 AM
OK, folks
What appears to be the reported problem here is the picture on the left. Nobody is reporting the picture on the right.

What do we know for certain? That the bit is travelling beyond the point we expect it to stop cutting then returning to the position we expected it to stop at.
What are the possible causes?
Something loose?
Software/controller glitch?
The carriage stops, but the router/spindle acts like a pendulum and swings forward then goes back to upright? I quite like this idea if it only happened in the y direction, but seems less likely in the x direction.
If it's too fast ramping, and the carriage actually over-runs at the stop, how does it know to go back to where it should be?
The carriage stopping in the right place, but the bit (z axis) continuing to move because of flex, then bouncing back seems a possibility.
It can not be incorrect toolpathing. The result would then be the right hand picture.
The more I think about it, the more the bouncing/flexing/pendulum swing seems to be the culprit. The idea that something is slightly loose, allowing the bit to 'grab' the wood and pull forwards, returning to its true position once the grabbed wood is cut away also seems feasible.
I guess tests with very slow ramping, very slow speeds, different materials should tell us something.
Today I disassemble my Shopbot for the move, so testing is some way from my mind


06-26-2006, 11:05 PM
It's got to be software related. I've not changed a thing in the way I toolpath my cabinet parts and I get this in both the X and Y. Today it happened in the middle of a 96 " cut(somewhere around 80" - 85") I've examined all my toolpathing, this happened with a new bit. Same kind I've always used @ 2.5 IPS. The only thing I have changed is the control software ( most recent update).Kip

06-27-2006, 03:40 AM
Which version of control software is everone who has this problem on? I have just put in 3.4.23 and havn't done anything yet that would show it up, but would like to test it if others are having problems with it.

06-27-2006, 03:47 AM
Also is it only an issue with PRT's and not Alpha's?

06-28-2006, 09:18 PM
After so many tests what i found out is that this phenomenon happening in plastic only.
No problem in wood, corners looks perfect in wood

06-29-2006, 09:18 AM
I've never cut any plastic. All my problems have been in wood.(HMM, lots of truth in that.) I'm running the most recent software, I forget the version #.

06-30-2006, 03:01 AM
Is it possible that you are melting the plastic as the drive ramps down to a stop before making the turn???

Check the cut with a magnifing glass or low power microscope. You might also try the same cut at lower RPM.