Oversize Laser Project (10W)

[liquidmettle]

Hi All,

I’m getting ready to embark on lasering a 3 foot mirror. I’m already planning to build some roller supports in front and behind the platform to take the weight of the overhang (thinking I’ll employ foam paint rollers), but I’m more worried about lining up the sections of the burn.

I made a small jig with 1-2-3 blocks to line up the edge of the mirror in the same place, but in a test burn using camera capture, when I positive some squares in a line below previously burnt squares, they were 2-3 mm off my expectation.

Should I just avoid using camera capture and measure everything to the nines using the coordinates? Is there a better way you can think of to ensure camera accuracy?

Any pitfalls I might need to be aware of in your mind with this project?

Thanks!

[sdj544]

Abandon using the camera.
Just not accurate enough.

-S

[liquidmettle]

I had fear this would be the advice. I’ve had good results with CC, but it doesn’t seem repeatable, even with a relatively flat stock like a mirror that should have virtually no parallax.

[Skreelink]

Don’t use the camera capture, it’s always been iffy at best. I have a guide on finding a repeatable true origin here:

Accurate, Repeatable Laser Guide

My rotary guide (found here) seemed to be well received, so I’m going to move onto the laser. This guide will be a bit shorter, as it’s not too involved, but is mostly to help increase repeatability, accuracy, and quality of your laser projects. Likely any machinist will say a repeatable, accurate origin is a key component in quality production. You gotta know exactly where you start to know where you’re gonna end right? I know the snapmaker method is just eyeball it, be it with a mark and lini…

Theoretically, if you can move and measure accurately, once you have your origin, shouldn’t be that hard to do.

[liquidmettle]

So following back up on this after a brief hiatus for other projects and general life interruptions.

I started by looking over your guide @Skreelink and I think its a great thread you put together. But I either jumped the gun, or didnt interpret one of your steps correctly.

II tried to start a burn from Luban (which may have been why my first issue happened). After securing the plates as you outlined, I put a small dot over 0,0 in Luban. I measured the material thickness (which came to roughly 10.7mm).

When I told Luban to start the job, and input the 10.7mm height, the laser head slowly lowered until it crashed gently into the bed and then retracted a moment later (thankfully my material was cardboard and no damage appears to have resulted to either the head or the bed).

I will try the more explicit manual jog next, but I am a little paranoid about a repeat of the crash without the repeat luck of the unscathed.

What do you think of the following process (worth pursing further or no):

I modeled some platform spacers that were open on the ends so that I could align one edge of my mirror consistently but allowing the full 35-36 inches to hang off either side as needed (the far right one).

I affixed a piece of carboard the same width as the mirror against the spacer and taped it in place. Then I used camera capture after putting a grid pattern into Luban and ran the job:
Grid:

Outcome:

I turned the machine off to resume the next day (today). I loaded the same test grid file and ran camera capture again, with this result:

You can clearly see that the lines/numbers don’t match up in Luban to the previous day’s burn. I added two crosses (40mm lines) at coordinates 30,50 and 290,50 in Luban.

Running a burn of just those two crosses resulted in:

The lines perfectly align with my 30mm and 50 mm lines from the previous day.

This makes me think I can simply determine the exact coordinates where I should move the mirror between burns so that the edge of the next section of picture lines up with specific coordinates. I plan to use the bottom edge of the spacers at the top of the platform (my spaces are 50mm long) as the line where I should move the mirror. Then, I can still use Camera capture and Luban to take advantage of the auto-height feature (unless this is available through the touchscreen as well, I haven’t tried that route since getting the 10W).

It won’t matter if the captured image aligns so long as I use whatever coordinates I determine to be the Y line to align each new section of picture. And since I plan to use the bed clip edge as the real world line to reposition the mirror, I’m thinking this will be a solid approach.

Would love your feedback if you see any flaws in my thought processes or approach. I certainly could improve the design of my spacers but wanted to do something simple to start.

Also, to ensure there isn’t pressure on the 2 feet of mirror hanging off the platform, I designed and printed the following to line up with the top of the spacers:

These were the first version, and I had to make them about 10 mm shorter after switching back to the laser bed for more accurate measurements, but I am pumped that version 2 seems perfect and to be a win for now. Those are 4-inch paint rollers (cut off one end to expose the bore) on a printed spindle and stand.

[Skreelink]

The plan sounds solid. Using the camera capture to just figure out the proper offset, then ignoring the actual results and using your calculated one is a good strat. The models you made to ensure you have a repeatable position is excellent. If you know exactly where you start, you know where you end.

Also, good thinking on the rollers, overhang pressure would bow the surface and put it either out of focus, or out of balance and cause issues. As long as you have a good method of moving it forward the exact amount that jives with your offset, I think you’re good. (I will admit, I have been drinking a bit before reading your reply and making my analysis, but at the same time, a lot of my best work has come from having a couple and just seeing what I can make). Overall, good setup I’d say.

[liquidmettle]

I appreciate your very inspired () feedback haha

Cheers!