Laser engraving - not really getting great results

I think I’m missing something in my learning path with the laser. I thought I’d try some engraving (simple plague with some swirly things and text). I tried Lightburn - that’s the real scribbly stuff second up from the bottom:

[attachment file=87055]

Then I tried Image2Code - the “not MPCNC friendly version”. That’s the bottom most one in the pic above.

And then I tried the MPCNC friendly version. That’s the top attempts here:[attachment file=87056]
[attachment file=87057]

I know there are setting you all will need to know in prefer to help me but not sure which ones you’ll need. Just let me know and I’ll back track my brain. I can also post the gcode if you think that’ll help you see what I’ve got not right here. I appreciate the help.

To me it seems like the laser wasn’t firing at all the spots where I thought it should be firing. Is it possible it just flashed so fast I didn’t see it? The Lightburn stuff seemed to only do the outlines but I had it set to “scan”. Perhaps loose belts on the machine also making it just imprecise enough to mess things up?

I haven’t tried that program yet, I’m using the Inkscape plugin to start off with. It seems to be pretty easy to start off with, to get the laser focused right. Here is my first try on wood.

Yeah we still haven’t figured out the path of least learning on the laser software. I use the inkscape plugin for vector stuff as well. I think Aaryn is really making progress with the laser stuff in his thread, and Victor just updated his software as well.

I’m going to set up the speeds and power shader that he posted and see if it helps me figure anything out. I’ve tried two Inage2Code photo etches and they were both just smudges.

Line art is no problem at all. It’s the shading that’s eluding me I think.

Hi Kelly, we were talking about the shading programs in this thread.

 

Victor is doing some great programming for us. I said I would test a few changes he made to his program and the laser Grbl app. But I haven’t tested them yet because my wife asked me to 3D carve a few items. I’ll switch the machine back to laser soon and test them.

1 Like

Thanks Aaryn. That’s a great thread. And thanks for the link to the speed/power thing. Saved me creating one from scratch. I’ll keep digging through but wanted to share this image of the test. Looks great to my eyes for a virgin setup and definitely gives me a great stepping off point for troubleshooting and tweaking.

[attachment file=87239]

1 Like

Okay. So I’m still a long way from where I want to get but I think I’m heading in the right direction. The bottom band is a file with the feedrate and travel speeds both set to 12mm/s. I thought maybe if I sped it up a bit it would lighten so I made the second test there with both speeds set to 15mm/s. I didn’t see a noticeable lightening and actually saw some noise coming into that light area to the right. Also, the machine was stuttering during all of the transits on the x axis. Like it was slowing down to burn even though the speeds were set to match. I went back to the 12mm/s file and just let it run so I could see what sort of luck I had with the details. Kinda streaky…

So it’s way over burned in general. That’s one issue. It’s also streaky. I left the min max in Image2gcode set at 0 and 255, chose “sharp” instead of smooth and resolution at .1mm (could this be one thing I need to change? I have no idea what size ‘dot’ I get from the laser)

I noticed that the latest version (just published last month by Vicktor) doesn’t have the contrast adjuster like the ‘modded for MPCNC’ version. However when using that version (the modded one) all I got was a laser-off perimeter move and then no laser at all as it started the engraving paths.

I’ve decided I need a better and faster to run test file so I’ve created one and it’s running right now. I’ll post it before bed here. In the meantime here’s these pics…source photo included (I know it’s not the best in terms of contrast…but I would think it should make a nice engraving still no?). Oh, and the ‘oops’…I bumped the plywood moving a fan around in the enclosure right near the end of the run. D’oh!

[attachment file=87276]
[attachment file=87277]
[attachment file=87278]
[attachment file=87279]

Pretty dark. I’ll play with this at different speeds and see what I get. This was 12mm/s again. Source image here as well. Those are 5% increments in black, starting top left and going back and forth like a snake to the bottom left corner at 100% black. Or vice versa. The pic flipped when it posted… [attachment file=87285]

[attachment file=87284]

E6CCD618-F417-4CC1-B3F8-BC915DA591C2.jpeg

Just wanted to add…I thought maybe I should try wiping that off in case there was a bunch of soot on the print and sure enough there was. The gradation was way more visible than without having rubbed it off. How is everyone clearing that off the print? Does the “air assist” help with that? I’m firing up the compressor to blow off the next two tests (15mm/s and 20 mm/s).

Before the rub:[attachment file=87291]

Something still is not right with that.

Running one now at 10mm/s but with the max power slider down to 200. I’m assuming that 50% power would then be 100 as opposed to 127.5 if it were left at 255. It’s about halfway through.

The 15mm/s gave better gradation but got pretty streaky and the vertical edges were not sharp at all. The 20mm/s was pretty messy. I thought maybe my belts were causing the jitters but it doesn’t appear that way. I just watched one of David’s old videos from 2016 on a cardboard pizza box and am nowhere near those results. And he said he was running at 50mm/s!!

Here they are in all their glory. Not sure why the perimeter line is so jaggy. I wonder if it’s because of the grain of the wood which is running perpendicular to the vertical lines there? At “full” power there’s some bleed in the burn. I’ll double check the focus.

[attachment file=87312]
[attachment file=87313]
[attachment file=87314]
[attachment file=87315]

I am curious. is your laser lens firmly in place? I noticed on mine it is loose where I can turn to focus it. When I touch the focus it will move the beam about a millimeter. I was worried it might be loose enough to cause the beam to jitter as it moves. I haven’t seen it happen yet on my machine but I would guess these pictures are similar to what it would look like if the lens was bouncing around. Maybe not. It only seems off in the X direction.

Oh wait. Are you using M03 or M106 for your laser control?

Interesting- I’ll check that in the morning and let you know. I’m using 106/107.

Alright… I hadn’t been following this thread too closely but I noticed KellyD’s reference to one of my videos above… and decided I’d better chime in. But, first, I had to dust off my MPCNC “Joe” laser engraver that has languished for some months…

I decided to see if I could duplicate my results from a couple of years ago, so I found and downloaded that popular Garfield image once again and processed in with the MPCNC-custom Image2Gcode last loaded on my silly Windows7 box. I used pretty much all the defaults and the “Wood” profile… which sets in 35% max power. It defaulted also to 3000mm/min travel/feed so I went with it. Other than being a bit light, the results didn’t seem too bad for an initial run…

[attachment file=87384]

Then I decided to visit Victor’s Thingiverse/ImageToGcode page and saw that our own Jeffeb3 had piped up there with compliments about Victor’s software and a note that he’d gotten it to run under Wine. Well… not to be outdone, I downloaded the software and managed to get it running under PlayOnLinix/Wine as well. Again, not knowing what I am doing, I went pretty much with the default power 0-255) and set in 51mm/s (3060 mm/min) as both travel and feed rates. Though possibly a bit on the dark side, I was thrilled with the result…

[attachment file=87385]

KellyD, I think you are running way slower than you need to… if you are getting full power from your laser (which I think you are, based on the charring you’re seeing…). There may be other things going on with your machine as well IDK… but ASSUMING that my machine is anywhere near accurate I’ll probably need to reduce my power and/or speed up my travel/feed rates. But I am truly impressed with Victor’s ImageToGcode converter – best I’ve seen yet – and it runs on my pretty-stock Marlin-based MPCNC!

– David

 

 

 

 

2 Likes

So grateful for your time in digging in like that. I think there are a couple of things I can check into as far as the machine is concerned. 1, the lens as Aaryn suggested and 2, the belts and bearing/rail tensions. I’ll also try running angled instead of horizontal - maybe if both axis are active I’ll have better results? Garfield has been downloaded (might as well keep apples apples) and I’m going to get a test running while I work. I’ll run on cardboard too - oranges can also be oranges.

Did you use a greyscale image or any special prep before taking into ImagetoGcode? I did a bunch of reading this morning and a lot of the current info says to convert to bitmap first. Some even say varnish the wood first…

I tried the “shades of gray” image that you’ve been using… interesting results. I used Victor’s ImageToGcode again… 0-255 on the power levels and 50 mm/s for both travel and feed. This was done on cereal-box cardboard (my favorite, cheap, test material) and the lightest shades seem to almost have a “bleached” look to them. Also, note that the right and left edges of each block exhibits the effect of acceleration/deceleration, I think… leaving the edges a little ragged and darker due to the longer dwell time for the laser.

[attachment file=87392]

[attachment file=87393]

I think our MPCNC’s and lasers are set up similarly so I’ve attached (I hope!) the actual gcode files from ImageToGcode for the smaller Garfield and the “shades of gray” chart. I didn’t do anything to the images beforehand… they are simple JPG images (also attached).

[attachment file=87398]

[attachment file=87391]

– David

imagetogcode.zip (734 KB)

garfield-shades_of_gray.zip (33.3 KB)

3 Likes

Wow, the accelerations are clearly visible, that would be pretty cool if it didn’t totally mess up the end result.

That’s because marlin doesn’t care about adjusting laser power according current speed.