Engraving shifts but the object is not moving!!! whats going on?

[Mohammed Issa]

hey guys,

this happened twice before i decided to tape down the mirror, i thought the air assist was shifting the mirror around. but after taping the mirror down, i found that the machine itself is doing something wrong.

this is my chinese machine. its running on 400mm/s.

the belts are all fine, the tightness is perfect.

anyone has any idea whats going on?

thanks!
photo.jpg

[Rodne Gold]

Can't quite see whats going on in the pic but maybe the steppers are missing steps for some reason , try reduce the speed to around 250 mm and see if it helps

[Mohammed Issa]

it doesnt happen all the time, it just happens every once in a while. it happened about a month ago, i didnt know what the reason was, i just reset the machine and it never happend again until yesterday. at first, i thought the mirror was just floating around. i didnt care much. but when i went back to trying again today, it happened twice (not consecutively). so i reset the machine, but it didnt happen again yet. (although i reset the machine yesterday as well).

thanks,

[Walt Langhans]

In RDCAM under the Config tab is a box that says Export Setting, the default is 80%. I had my laser do some odd things but when I switched that to 100% every worked fine so I would check that.

[Richard Rumancik]

Mohammed, I see that in the picture all of a sudden the image shifted right and plotted several lines like that; then it shifted back. You say that you checked the belts but did you check that all the timing belt pulleys are tight to the shaft? It just make me think that perhaps a pulley slipped some number of degrees and then after a number of strokes slipped back to the original position. So I would suggest that you check the setscrews.

Edit: it is hard to know what is being shown in the pic. I was looking at the center piece as an isolated plot. If my theory is correct then it would occur across the whole horizontal (x) sweep, not just in one section of it. I see that the image section to the left is not shifted so if it was all plotted at the same time my theory won't hold.

If it the shift is occuring in just one section of a complex image then it is not mechanical.

[Mohammed Issa]

here is the one that happened yesterday, this might help clarify what the problem is.

its just a scrap mirror, so dont mind the holes in it.

its supposed to be a rectangular box around the photo (its a bad photo engraving trial, i was just testing things out before going to the actual mirror).

so you can see the whole box gradually shifted to the left in the middle of the engraving. this was a BMP file.

image2.jpg

[Mohammed Issa]

alright, so here is whats going on.

its 100% not hardware.

i was working on the actual mirror, and it actually did it on it as well! so, i wanted to see if it was the file that was sent the machine, so what i did is i stopped the machine before finishing the whole picture. then i pressed START again, that way it will use the same file that was sent to it. and it actually made the same mistake again, in the exact same place.

so aparently, the file is being sent like that.

i actually went with Walt's suggestion, (Export Settings >> Export Precision 100%) before sending the file to the machine, so that didnt solve the problem.

any other suggestions???

thanks!

[Richard Rumancik]

This error is quite different than the other one. Pretty puzzling. Have you tried to swap out memory in the laser? It is not necessarily that the computer is sending a bad file; it could be getting corrupted when mapped to laser memory.

Can you print the file to a paper printer? If the file will print on paper correctly, that should rule out the PC side.

[Dee Gallo]

The only time I ever had something like this happen, it was the side mirror (the one you don't see easily). Once I cleaned it, everything was good. Now, every time I clean the encoder strip and lens, I clean all the mirrors as well. Dust on the mirror moves around, so your effect can too.

[Mohammed Issa]

The only time I ever had something like this happen, it was the side mirror (the one you don't see easily). Once I cleaned it, everything was good. Now, every time I clean the encoder strip and lens, I clean all the mirrors as well. Dust on the mirror moves around, so your effect can too.

hey Dee,

that could happen. but what happens is when i ask the laser to engrave the last file, it make the exact same error again, so it has to do something with the data thats being engraved. somewhere along the way, the x and y coordinates are getting corrupted. my problem is, i have no idea where/why/how to solve the problem.

thanks,

[Dan Hintz]

My guess? This is one of two things:
1) Electrical issue... poor grounding or some such is making the stepper drivers freak out and gain/lose steps.
2) (more likely) print driver issue... if the same data set produces the same issue, that data set is tripping over a bug in the driver.

[Scott Shepherd]

hey Dee,

that could happen. but what happens is when i ask the laser to engrave the last file, it make the exact same error again, so it has to do something with the data thats being engraved. somewhere along the way, the x and y coordinates are getting corrupted. my problem is, i have no idea where/why/how to solve the problem.

thanks,

Not saying this is the issue by any means, but just to clarify your logic- running the same job twice doesn't prove it's not mechanical by any means. For instance, if you have a bad spot in the 1st or 2nd mirror (dirt, etc.), it can work fine until it hits a certain place on the table and then that beam is hitting the bad spot, it shifts the beam over for that area, and you end up with the shift. You can run the same file again and get the same result. Trust me, been there, done that, but nothing to the degree you're looking at.

So don't rule out a mirror issue. And depending on the size and placement of the job, you might not hit that bad spot while lasering other jobs. I'd do a double check on ALL the mirrors and make sure they are all clean and not damaged.

[Richard Rumancik]

Mohammed - do you have to stop (abort) the plot to be able to repeat a plot? I would have thought that you could let it finish and repeat - that's how most lasers work.

Do you have the option in the driver of setting the laser carriage so it does not "home" after a plot? Then when it make an error, move the carriage to a much different location on the table and repeat the plot there. If it repeats the error in the other location then you can probably rule out the bad mirror theory.

In the first example (photo.jpg) there was a slice of the image that was offset and I suggested that maybe the drive pulley was loose. The second image (image2.jpg) looks quite different in that the offset occurred gradually. In thinking about it a loose pulley could do this too if it was nudged a fraction a degree on each stroke, and then eventually come to a hard stop when the pulley locks up again. I know you are 100% convinced it is not hardware but I'll play devil's advocate here, so humor me . . .

I'd be interested in whether your machine is maintaining the coordinate system after an error occurs. For example, in image2.jpg, the plot starts in the top left at (x1,y1). After the error, the bottom left corner of the image under the hole is at (x2, y2). The question is: does your machine think x2=x1? The answer might shed some light on where the problem is.