Square, uh - not square?

[silverlaker]

This is my first post - however have enjoyed using this forum as a resource. I have a 6 month old prs alpha that has work perfectly, save for user error. Now, however I have something happening that I can not understand. My machine is not cutting square in one quadrant of the table (closest to the origin point). If I set the origin 40 inches down the long axis - I get a square cut. If I cut a rectangle the size of the table, I get a square cut. If I cut a 2' square in the quadrant closest to origin point I am consistently 1/16 out by pulling diagonal measurements. Here is some of what I have tried to resolve the problem:
update the control software, restart computer, change bits, collets and collet nuts, adjust the racks, tighten every bold on the machine. I have placed a six foot level and /or a straight edge anywhere they fit to ensure that the table is level, that the rails are not warped, and that the gantry is not twisted or warped. I have resurfaced the table and had a beer at lunch. Im really hoping for a simple user error - but I'll take any suggestions you have to offer!

Thanks in advance,

Tom

[myxpykalix]

Have you taken a measuring tape and measured the table diagonally from one corner to the next to make sure it is square on both sides?

Also can you cut a 2' square in another place on the table that you know will be square? If so, then cut a square making note of north,south,east,west like in the diagram in black.

Then try to cut one on the bad part of the table making note of north,south,east,west then compare them to see which part is the affected side. This might help you narrow down wheter it is a rail or carriage problem.




3897

[dana_swift]

Thomas: three suggestions:

Clean the V rails, dust and grit accumulating on the V rails can cause the car to shift in that region of the track. Check the rack and pinion gears for being clean. Especially the rack.

Cut the same place with a climb vs conventional cut and see if it reveals slop or backlash somewhere.

When the bit is located over the trouble area, manually push on it in X, Y, and Z directions and see if there is any slop there. Compare to the good region of the table.

Good luck-

D

[silverlaker]

Thank you Jack and Dana for your thoughts. Dana, I followed your suggestions and am happy I did, as the the machine needed a good cleaning.I cut in opposite directions and checked for irregular movement. Alas, there were no inconsistencies and the bot is still cutting 1/16" out only in that particular quadrant.

Jack, following your advice led me to conclude that the y axis is not cutting square to the x axis in that particular area. This is very strange. If there were a problem with the movement of the z car it would show-up across the entire table. I have always suspected some issue with the way the gantry is rolling, a twist in the rails, non parallel movement, something... But I cant find anything. Also I did pull diagonals on the table and it is perfectly square. I still appreciate any thoughts anyone has to offer - Im sure Im missing something. to be clear about the issue, please see attached diagram.

Thanks -

Tom

diagram
square not square.pdf (http://www.talkshopbot.com/forum/messages/26/square_not_square-49189.pdf) (12.1 k)

[jerry_stanek]

I would try and switch the rails from one sid to the other and see what happens. one rack may be out of specs and bindeind the y across the table

[curtiss]

Are your rails perfectly parallel ?

Seems like the y-car is binding up into a parallelogram at the problem end.

When I was squaring up, I made a "long adjustable feeler gauge" with a 4' ruler and some 3/8 threaded rod + gorilla tape.

It helped to make sure the rails were parallel for full length.

[rb99]

Have you checked to see if the rack is clean? Maybe it is clogged up and causing it to skip a gear or something?

RIB

[tparr]

It might be helpful to either disconnect the motors from the racks or electrically and pull the "X" & "Y" through the complete motions to see if you can feel what the machine is experiencing.
Just a thought.........

[chiloquinruss]

I found I had to put some very thin shims under one section of one rail. As I was watching the carriage roll down one rail I noticed that it did not touch competely both sides of the v wheel. It was tracking straight but just not quite touching solidly. I put in the very thin shims and retried the test. It rolled through smoothly. I then recut an old project on a piece of oak that I had cut a month before but it had given me some little chatter marks on one side of the long cuts. The marks on the recut piece were gone. My suspicion is that the not quite touncing v wheel was torqing over when cutting the oak but running true when running light. If you eyeball down the track with the carriage running very slowly you just might catch something similar on your machine.

[silverlaker]

Thank you all for your posts. I've had to focus on some production and design issues the last few days - but will be revisiting my CNC problem today. I'll investigate your suggestions and report back.

Thanks -

Tom

[dana_swift]

Thomas.. been thinking about your problem, are you sure it was correct at some point in the past? The real question is "did something change?"

Warped V rails, something becoming loose, etc. Somewhere this has a simple rational explanation.

Good luck.. wish you were local, I would love to check this one out!

D

[city_graphics]

Did you ever correct your problem?

I had a similar problem that still lingers in my world.
Over the summer while cutting a file with an endmil, in the middle of the job, the Z just dug down and the machine tried jogging home. The result was a broken bit, a ruined job, and the gantry racked and off its track.

I spent days trying to fix the problem. As I roll the gantry manually I see the Vee rollers showing daylight at the far end of the table right front and left rear.

I rehecked every possible thing on the machine: The square, the level as well as the distance between rails. All was fixed and every bolt retightened.

I ended up breaking all the gantry bolts loose and the gantry did settle but I still get some daylight at the far end. But that is not my big problem, when the gantry gets to the far end it must be climbing up because the Z does not cut as deep on the far end as the home end.

For me, something is racked and I have beat my brains out trouble shooting the problem.

And the kicker is, I CAN CUT A PERFECT SQUARE.