[Railmen]

https://youtu.be/XJTbiqtvzFUOk so you will have to bear with me, I new at this CNC stuff, I bought a used home built table.

software: Mach3, sheet cam, corel draw x7 home and student.....
hardware: I believe that they are gecko's, PMDX board THC300, linear rails.....4x4, windows XP

the main problem is that the gantry gets way out of sink on long runs down the table, not always but 80% of the time it does.... if I reference home the gantry takes off faster on one side then the other(Y and y slave).

sometimes cutting larger circles it might be out by .100-.125" then other times bang on.

I have a small video of it, will try to get it loaded some how..

sorry if I missed some info, just ask me for more and I'm sure I might be able to get it, I would really like to fix this, its getting old fast and lots of scrap

Thx Railmen.

[CaptainVee]

I'm no good at the electronics side of things but here are my first thoughts.

It looks the movement of the motors is more ore less independent.
If they get the same commands, they should either make the same movement or stay moving skewed if one motor loses steps.
I can not hear that either one does lose steps...

The other reason that might explain this is that the motors are closed loop (those want to correct it if they lose steps).
If they notice losing steps they might correct it in the way you see.

In that case there should be friction in the gantry on one site when there is non on the other.

The last thing I can think of is that there is a fundamental mistake in the configuration, or there is something broken in the machine.
If the motors are the same on all axis, you can swap the x or Z motor for one of the y motors and see what happens.

Oh, better still, do that with the drives first, that is easier.

[Torchhead]

1. Those are steppers so there is no position feedback
2. In MACH the slave axis MUST be tuned the same as the Master axis if they are the same type of motor
3. Most of the time if the tuning is right then its mechanical (check all of your couplings, and set screws on EVERYTHING
4. If it still does it there may be an intermittent connection on the one motor that causes it to stop and just vibrate.

The motors should move at exactly the same distance and in sync as long as the motor tuning is the same

[daniellyall]

if you wont to check if it squire clamp a sharpe to the torch lower it down so it just touches a piece of what ever draw 3, 4, 5 squire whats X 3inch or 300mm, Y 4 inch or 400mm then put in x0, y0 whats the correct angle to do 5 inch or 500mm if it goes back to x y 0 its squire if not its out.

to squire it if it does not have correctly set homing switches draw a line along X axis then another line along Y axis use a big squire, have it squire to the x axis line then where it sit out on the y axis line that's how far it needs to be moved to be squire

[Railmen]

1. Those are steppers so there is no position feedback
2. In MACH the slave axis MUST be tuned the same as the Master axis if they are the same type of motor
3. Most of the time if the tuning is right then its mechanical (check all of your couplings, and set screws on EVERYTHING
4. If it still does it there may be an intermittent connection on the one motor that causes it to stop and just vibrate.

The motors should move at exactly the same distance and in sync as long as the motor tuning is the same

K will Check that........ I think.........

it must be close cause it cuts most of the stuff fine, just seems to have a gremlin some where in it..........

seems like the slower the cut speeds its fine, but get it jacked up and off it goes.....

I was cutting some 1/16" alum and having the speed up around 230IPM and my square holes were rounded corners, thought that maybe my drawing was wrong(not like that ever happens) checked it and it was fine........ so just for fun, I slowed the feeds way down and low and behold the corners became square again.

so maybe my steppers are just to low on power to drive the gantry fast enough although Mach3 thinks its made those steps in realty it hasn't???

does that maybe sound possible???

Thx everyone for your help, Railmen

[daniellyall]

that's sound right the steeper`s or power supply may be to small, so if fast its bad slow its good could be steppers or power supply or drive settings.

easiest way to find out look at the size of the stepper`s, pulley`s, R an P or ball screws you have what moves the axis, work out what weight or force is needed to move each axis then work out from there if they are the wrong size.

there are some calculators on line to help work it out.

someone might post the calculations for you

[alan_3301]

One side of your gantry blasts off in that video. That is a lot of acceleration.
Were you homing the axis in that video? If yes you need to read up on all the homing parameters, and figure out what is wrong.
If not, you either have some crazy gcode, or some even crazier electrical interference going on.

As far as rounded corners on squares, I would read up on all the CV settings on mach3.
Sorry I cant be more detailed I don't use mach so I'm not familiar with those settings.

[LeeWay]

It looks like it is binding badly when homing. Then it breaks the torque on one side to relieve the bind and allow the other side to move again. I had a bearing go out a while back and it acted similarly. That was the only time that I ever saw the gantry hit the switches at different times.

[Torchhead]

While CV is the cause of the rounding, it's trigger is the fact you do not have enough acceleration to make that move at that speed. The reason it works better at slower speeds is it takes less time to accelerate from 0 to 100 IPM than it does 200 ipm.
CV is there to smooth out the motion otherwise it shakes like a dog crapping a peach seed. It stops and starts on every node. Now you can go in and dink with the CV acceleration and angles and it will improve on some cuts and worse on others
A better solution is to deal with the Physics involved. Simply put you need to tune for best acceleration on XYA and if its not enough wither use SheetCAM and a cut rule to slow down on corners (which BTW will cause your THC to dive) or consider higher torque motors or trimming some weight off the gantry (or both). It difficult to run a dumptruck in a road race with lots of turns

Acceleration is defined as Force (linear force on the load provided by rotarty torque from the motors) divided by the MASS (weight on earth/32) A = F/M Pair that with the fact stepprs lose torque the faster you spin them until they stall (lose steps)

If one side is under more friction than the other it will stall sooner If you only see the racking at higher speeds then consider that high speeds are part of the viable clues. Given what you now know about steppers and RPM and motors stalling under load use that it do some testing and to consider what to do to cure the issue. Some tables just cannot handle cutting at higher speeds and changing directions instantly.

[Railmen]

One side of your gantry blasts off in that video. That is a lot of acceleration.
Were you homing the axis in that video? If yes you need to read up on all the homing parameters, and figure out what is wrong.
If not, you either have some crazy gcode, or some even crazier electrical interference going on.

As far as rounded corners on squares, I would read up on all the CV settings on mach3.
Sorry I cant be more detailed I don't use mach so I'm not familiar with those settings.

Yes I was homing it, I know that it always gets out when I home it.......

The more and more I think about it, it sure seems to make sense that it's just under powered.....

I want to build a bigger table with the parts I have but I'm thinking that maybe to just buy all new parts to over build the next one.....

Once again thx everyone for your help, Railmen

[LeeWay]

I don't think it would be under powered. Rather not my first guess. The machine is moving slow when homing. It is using profile rails. It has geared steppers on it. All of those things typically allow for very small motors to be used. It is binding somehow whether electronic or mechanical.
Have you looked at the motor tuning page? What are the tuning parameters? Check that the X axis and the A axis are indeed tuned the same.
What type or model of Gecko drives are you using? What size power supply?