Layer alignment problem when doing multiples with Chinese laser w/laserworks.(2)

[Jeffrey Michael Wilson]

322356I have a 50 watt Chinese laser. Not long ago we had a power spike, took out all our florescent lights and multi strips, blew the fuse in my laser and I ended up putting in a new tube and power supply...it's like 50% better than when new. BUT, not long after replacing supply and tube i noticed when I did multiples on an X axis, the layers as it got away from the origin...mis-aligned. My wife suggested I try the Y axis, and that worked, for about 2 weeks...Now either way the layers become misaligned as it gets away from the origin...BUT in some cases the first and last multiple, will be aligned. Belts are tight....This makes no sense to me, it was a gradual decline.... What the Heck. It doesn't seem like a stepper motor thing because it is so consistent, and on longer multiples the first and last will be aligned. The picture shows one done on the X axis, and one done on both. There are 3 layers, the bee design (raster) and holes (vector) and cut out circle on the outside (raster). Do any of you Experts have any ideas about this?

[Bert Kemp]

Do you see the little hour glass spinning in all of our heads, I'm going out for a bike ride and ponder, and hopefully someone will figure this out while I'm gone, because i can't think of anything :(

[Rich Harman]

Which main board are you using?

I remember occasionally having similar problems, but not since I upgraded the main board/LCD to the RDLC320. I'm not sure that was the problem, could have been a coincidence.

[Jeffrey Michael Wilson]

Yeah, it's a major PAIN...If I have little tiny 20mm pieces, I have to do them one at a time...
Do you see the little hour glass spinning in all of our heads, I'm going out for a bike ride and ponder, and hopefully someone will figure this out while I'm gone, because i can't think of anything :(

[Kev Williams]

I don't know what the problem is, but I'm familiar with it. My Triumph has done that before when changing colors. I even made a video of it doing it. I never did figure out the issue other than you can see and hear the machine change speeds as it travels from the finished color to the next (didn't mention that in the video), even though the speed and power settings were identical. The resulting offset is just about equal to the time difference- in other words, the machine used the same amount of time to change, but because it sped up (in my case), it went farther away, missed by nearly 1/4". It almost acted like the X stepper motor changed steps-per-mm during the change. I tried changing colors, changing the order of the colors, that didn't help. I didn't spend a lot of time trying to find a cure other than avoiding many color changes, and testing before I run the job.

The really weird thing was, after it missed the position of the next color (cyan in this case, following blue), every other color was wrong too. BUT, after stopping the job and running ONLY the cyan, it would engrave it where it was supposed to go. Why that's weird is, it did this without resetting the home position, so the steppers didn't skip. It just simply engraved the cyan and following colors in the wrong place!

It's only happened on one particular job and layout, but it did it every time, and I'm sure it will do it again if I run the same job. (Large SS operator panel) I was going to contact Triumph but I've just chalked it up to a software or power supply glitch. >EDIT< Just remembered, I have installed updated software since then, I should give it a go and see what happens...

Have no clue if you're suffering the same problem. But for fun, I would suggest trying changing the orders of the colors, or maybe change up the speed settings a tiny bit between the colors and see if that helps

[Jeffrey Michael Wilson]

What is the expense for changing a mainboard, and what does it entail exactly..is it new software as well? I have RD CAM 2 ??? I think with laserworks software.
Which main board are you using?

I remember occasionally having similar problems, but not since I upgraded the main board/LCD to the RDLC320. I'm not sure that was the problem, could have been a coincidence.

[Jeffrey Michael Wilson]

I'll try messing around like that...been in a crunch for a while just needing to get things done. Another really weird thing it has done, just with some files is, it changes the size..it'll read out the correct size on the screen, and print twice as big...If I put a border around it and resize the border, but don't fire it...it'll be the right size.

[Jeffrey Michael Wilson]

It looks like I'm using MC-Board Mainboard.

[Rich Harman]

What is the expense for changing a mainboard, and what does it entail exactly..is it new software as well? I have RD CAM 2 ??? I think with laserworks software.

I'm not sure of the cost, didn't have to pay for mine. Changing it out is not difficult. You need to save your machine configuration so that you can upload it to the new mainboard and you will use a more current version of RDCam/LaserWorks with it as well.

[Dave Sheldrake]

backlash maybe?

[Rich Harman]

Also you need to eliminate the possibility that the material is shifting.

[Jeffrey Michael Wilson]

ThanksYeah, it's not a material shifting issue, checked that first.... I'll look into mainboards...is there one that you feel is best?
Also you need to eliminate the possibility that the material is shifting.

[Jeffrey Michael Wilson]

What does that mean?
backlash maybe?

[Rich Harman]

ThanksYeah, it's not a material shifting issue, checked that first.... I'll look into mainboards...is there one that you feel is best?

I only have experience with the RDLC320 and whatever the older one was. No complaints with the RDLC320.

[Dave Sheldrake]

What does that mean?

When the head changes direction it will take up the slack, that can lead to a creep in the placement of engravings Vs cutting

[Gary Hair]

When the head changes direction it will take up the slack, that can lead to a creep in the placement of engravings Vs cutting

It won't creep in the same direction, it will offset to one side or the other based on the direction it is going. If something is loose, pulley, belt, etc., then you would get creep, but not with backlash. My cnc mill has a backlash comp setting, you set it and it knows how much to over (or under) travel when changing direction, the only time it matters.

[Dave Sheldrake]

It won't creep in the same direction, it will offset to one side or the other based on the direction it is going. If something is loose, pulley, belt, etc., then you would get creep, but not with backlash. My cnc mill has a backlash comp setting, you set it and it knows how much to over (or under) travel when changing direction, the only time it matters.

Just an idea Gaz, last machines I had that had any kind of BL comps were Mazak lathes :)

[Jeffrey Michael Wilson]

Curiouser and Curiouser.. This should be really a good one for the "Sherlocks " out there. .SO I decided to try an experiment. Since another problem I had a while ago was that old files that worked perfectly..specifically for setting certain sized stones....had changed drastically, almost doubled in size, although they read the correct size, the output was about twice as big. I found if I drew a vector circle around my selection, and "selected all", then resized it to the correct size, then didn't fire the circle I put around it...it would fire the new correct size. Basically old files were not firing the correct size, even when the size displayed was correct. Resizing the whole thing with a NEW border corrected the problem. SO, I decided to make entirely new files...I also wanted to see if all the layers were vector...if there would be a problem...so I created a simple 3 layer file in the laser software, one that would show any "Shifting" of layers. I made a multiple of it and fired it...PERFECT.. I did it on x axis...perfect, Y axis..perfect then did 5 going up and down and sideways...an array of 25...WORKED PERFECTLY... HMMM??? So I changed one of the vector layers to raster, as this is my usual routine, I had 2 vector and 1 raster..... It was Perfect. HM??? So I opened one of my old files that I had a problem with...... did it both ways, X and Y multiples...it worked almost perfect, except the outer circle's on some didn't quite meet perfectly... So I tried another....PERFECT. So after weeks where every time I tried a multiple of even just 2, and got major slippage....now it's fine.....
We've been gone to a show a few days and the machine has been idle, and cooler than usual. I wonder if usage will make it start slipping again, in the meantime....What the Hell???!!!!

[Bill George]

I was going to suggest binding in the carriage or other moving parts, Dave is more or less on the same page.

[Jeffrey Michael Wilson]

So last night I went to do another old file...it fires off to the right...Very consistently..the same amount of distance..and probably a good 10 mm's off to the side...So now what I've seen is that my older files are not firing correctly. With it perfectly lined up and focused, it will begin firing offset...and others it will fire twice the size listed....and these used to work perfectly..Then if I re do the file from scratch, same settings and everything....it works perfectly...no offset firing, no larger fire ...About the binding the carriage, or slippage from movement. I don't see this as the problem because in an array of maybe 30 pieces, each time I fire the same ones are exactly off the same amount, it's very precise about which individual ones and how much, and consistent on multiple firings. I mean why would the firing size be doubled almost on my old files, and some fire offset to the side. New files are not the problem, and these files are not stotred on the machine?

[Kev Williams]

With that many weird issues going on (hich doesn't seem mechanical to me), I think your controller is bonkers. Or your computer is feeding your controller bonkers data--