Mach did again!

[ranchak]

Here we go again. This is the third computer and the same problem still comes up. I am running a program that drills some holes, cuts a few pockets and then profiles. This is the first time I ran this program and I noticed that I did not make the first pocket large enough, so I stopped the run, went to zero and made a few programming changes. The program ran fine then all of a sudden the toolpath runs off to wherever. Again the origin was reset or changed somewhere, somehow by Mach. Take a look at the photos, I pushed Go To Zero which was no where close to being correct. I manually ran the table back to where about 0,0 SHOULD BE, it is off by over 3.5 inches. Seems weird how three different computers are having this same issue. There has to be a glitch or bug somewhere. Also this is the second time that I ran that pocket program, the first time I had no problem. So I don't think that the problem is in the program. I ran the machine for about 10 hours today with no trouble. Any thoughts? I can't belive that I am the only one that is experiencing this kind of problems on a continuing basis. If the DROs didn't change I maybe blame the drives or steppers. If this hadn't happened on three different computers I would blame the computer, but the only thing that is consistent is Mach and me. I'm pretty sure it's not me

[ostie01]

When using run from here, I always use single block and TO GO distance, I do not trust run from here either.

If the distance from stock is say .500 inch and DRO show 2.000 inches, I know that there is a problem.

TO GO button is my best friend.


Jeff

[Hood]

The thing here is Mach is reporting it is at the correct position and shows the toolpath to be in the correct position. From my understanding Mach updates the DROs from its pulse output at the parallel port so if Mach was sending out the wrong pulses the DROs should show that it is way off.
 Is it always the same axis that goes off? I thought it was the Y previously?
Have you swapped hardware around recently, ie have you swapped X and Y drives or swapped outputs on the BOB?
Have you monitored the voltage of your power supply whilst cutting? it may be falling short when doing multi axis moves.
Hood

[Hood]

When using run from here, I always use single block and TO GO distance, I do not trust run from here either.

If the distance from stock is say .500 inch and DRO show 2.000 inches, I know that there is a problem.

TO GO button is my best friend.


Jeff

I dont see how RFH could be thought of as the problem as it started back correctly and after a while things started going wrong. If it had gone wrong right away after the RFH then yes it could be a problem with it.

Hood
Hood

[rcaffin]

I am running a program that drills some holes, cuts a few pockets and then profiles. This is the first time I ran this program and I noticed that I did not make the first pocket large enough, so I stopped the run, went to zero and made a few programming changes.

'stopped the run' - did you hit 'feed hold' first?
I have noticed Mach losing registration a few times if I just hit Stop, and the manual (somewhere) warns about it.

Cheers

[TramAlot]

have you tried a differnt controler, mabe the cooling fan took a dump

[ranchak]

I ended the program run, went Go To Zero, Closed G code and the Load G code (once fixed) and pushed Start.

When using run from here, I always use single block and TO GO distance, I do not trust run from here either.

If the distance from stock is say .500 inch and DRO show 2.000 inches, I know that there is a problem.

TO GO button is my best friend.


Jeff

I dont see how RFH could be thought of as the problem as it started back correctly and after a while things started going wrong. If it had gone wrong right away after the RFH then yes it could be a problem with it.

Hood
Hood

[ranchak]

It was the Y axis before that was off in tolerance. BUT both axis 0,0 would appear to be reset.

The DRO does not change, what changes is 0,0. That is the only answer (or so I think), how can the machine be off when you Go To Zero? Something is changing 0,0. The DROs are really only showing a "location", they do not show true distance. The DROs have no way of knowing where they are in relation to 0,0. If they did the machine would not run off. If 0,0 didn't change Mach wouldn't be trying to move to the "wrong" place.

I think from what I am hearing that this is the same or similiar problem that Terry had with a plasma cutter.

[docltf]

ranchak

have you tried running your program without any G00 moves in the code.
open your code with the text editor and change all of your g00 moves to g01 moves with
a speed a little faster compared to the feedrates in the code.this will help stable the code
while looking for problems.

bill

[Hood]

No its not the same problem Terry had or at least from what I understand of it.
As far as I know Terry would see the DRO showing the true position of the axis and also the toolpath would show the true position but it was not a position that it was meant to be, in other words when an axis shot off it showed that movement in both the DRO and the ToolPath.

What you are seeing is the DRO and Toolpath showing where the axis is supposed to be but the axis is not there. In other words it seems like your drives are sending the motor somewhere else other than Mach is sending it.
Why I dont know, it could be faulty drives or much more likely is your Step signals are getting noise. It could be Mach but as the DRO and ToolPath seem to show the axis should be other than where it is I cant see how.
Hood