Anodized Aluminum Over-Engraving on Scan-Direction Lines

[Mikah Barnett]

Hi guys, I just joined but have been lurking here for nearly a year. I bought a G.Weike LC1290 in January for my business and have been having great success with it so far. I have one issue when engraving Anodized Aluminum (which I do a lot of) that I've yet to find a good solution for. I've seen one other thread here about it but no real solution, so thought I'd try again.

You can see in the attached picture (an etch I did on a rare scrap part) that when lines of an image are parallel to the scan direction of the laser, I get more etching than on the lines that are not. I have played around with many different settings but this seems to be there regardless of what I do - from low-power multi-pass etching to high-power one-pass and everything in between.

Has anyone else experienced this and found a solution? The one thing I have not tried is changing the logo file itself from 100% black to being a dark gray on the horizontal lines and fading into black. I suspect this might solve the issue on my logo, but then I'd have to go through a trial/error process with every logo I do. If that's the only answer, so be it. But, if there's something else I can try or change to avoid that - that would be very much preferred.

Can I just re-run or power-up until all the lines look like the horizontal ones, or will I over-etch and ruin the anodize?

267276

[Mike Null]

I think you need to reduce your power as you are removing the anodic coating on the horizontal lines. I'm sorry but I can't be of more help with your question.

[Matt Turner (physics)]

I'd guess that this might happen because of changes in the power of the laser depending on how long it's firing. On the horizontal lines, the laser is on for longer, and its power might increase because of that. You could see if this is the case by doing something like a dithered 50% at high DPI and see if this still happens. I'm not familiar with the Weike software, but there might be a setting somewhere to adjust the compensation for this behavior.

[Mikah Barnett]

Thanks for the input guys - both helpful. I share your thoughts, Matt, that it has something to do with "spool up" time of the laser. I can't find any settings that seem to indicate an ability to change that, but I'll keep looking. I think on the next run I'll just plan to do two passes at lower power to see what I get.

[Dave Sheldrake]

It can be changed Mikah in the software settings but remember DC tubes are non polarised so will often pump more (effective) power in one axis than the other. (not an exact explanation but close enough)

best wishes

Dave

[Rich Harman]

...but remember DC tubes are non polarised so will often pump more (effective) power in one axis than the other. (not an exact explanation but close enough)


??? I really doubt that the tube being DC has anything at all to do with a difference in power relating to the axis of travel. Please explain.

[Dave Sheldrake]

Randomly polarized light is exactly what it says; the light is plane polarized, but the direction is unknown, and may vary with time. Random polarization causes problems in optical systems since some components are polarization sensitive. If the polarization state changes with time, then the components’ transmission, reflection, and/or absorption characteristics will also vary with time.

best wishes

Dave

[Mikah Barnett]

Dave - care to give me a hint where in the software? I am using LaserCut 5.3.


It can be changed Mikah in the software settings but remember DC tubes are non polarised so will often pump more (effective) power in one axis than the other. (not an exact explanation but close enough)

best wishes

Dave

[Dave Sheldrake]

File/Machine-options/Work Mode(Engrave)

You can change most of the acceleration figures and distances in there,

best wishes

Dave

[Mikah Barnett]

Okay, that was the only area I saw that I thought might be it. But, since the problem seems related not to head acceleration but laser power on/off, I did not anticipate those settings making any difference. Seems to me (I could well be wrong) that during both the "intermittent" on/off periods where it's engraving vertical/diagonal lines and the steady-on periods where it's doing horizontal lines, all the acceleration should be done and it the head should be moving at constant speed.

Am I off base?


File/Machine-options/Work Mode(Engrave)

You can change most of the acceleration figures and distances in there,

best wishes

Dave

[Dave Sheldrake]

Okay, that was the only area I saw that I thought might be it. But, since the problem seems related not to head acceleration but laser power on/off, I did not anticipate those settings making any difference. Seems to me (I could well be wrong) that during both the "intermittent" on/off periods where it's engraving vertical/diagonal lines and the steady-on periods where it's doing horizontal lines, all the acceleration should be done and it the head should be moving at constant speed.

Am I off base?

Not off base at all but the rise times of Chinese tubes aren't great (to say the least) so the linear distance the head travels in a given time period will affect the amount of power introduced into the workpiece. You may find a shorter acceleration distance helps (so by the time the tube fires on the intermittent) the power has stabilised (power rise / time). Be aware though that setting the accl distance too short may put too much load on the steppers leading to early failure (it will also depend on which of the Leetro cards you have installed)

best wishes

Dave

[Dan Hintz]

I'd guess that this might happen because of changes in the power of the laser depending on how long it's firing. On the horizontal lines, the laser is on for longer, and its power might increase because of that.

Matt has hit it. Reducing the power will get you a nice engraving on long horizontal lines, but it may be too little power for a decent bleaching on the short lines. If your system has a bias setting (a.k.a., "tickle" in Western machines), increase it a bit... this should allow the laser to fire more quickly for short lines. Be careful, however, as too much bias will lead to lasing in areas that shouldn't be touched.

Polarization of the beam really only has an effect when you get up to very high-power machines (kW level)... think of it in terms of a short line rather than a dot. Moving in the direction of the line (like ---------- in a horizontal scan) leads to a higher average power per unit area than if the same beam was scanned vertically.

[Rich Harman]

I don't think that increasing acceleration is going to help, the acceleration happens before it starts firing.

You might try lower power and slower speed rather than lower power with multiple passes, if that doesn't work then widen the vertical lines - that's what I do.

[Dee Gallo]

Isn't this one of those times when you could rotate the image so there are no perfectly horizontal lines? This would avoid the situation of the scan direction lining up with your art.Or am I missing something important?

[Balsanu Gabi]

Hi,

I have that problem - different thickness between horizontal lines and vertical lines. It end up being the tube was not aligned properly. It was a little bit up on the head.
Check that. Just an opinion from experience. It will be nice if manufacturers come with a bullet proof system to align the tube properly. On a CNC machine you have many mechanical parts moving. On a laser....


HTH
Gabi

[Dan Hintz]

Isn't this one of those times when you could rotate the image so there are no perfectly horizontal lines? This would avoid the situation of the scan direction lining up with your art.Or am I missing something important?

I considered that suggestion as a possible solution, but the problem is it can lead to jaggies on diagonal lines when the resolution is set low (which you often do for speed of processing). Plus, it would make the engraving box larger, which means more time spent going back and forth (i.e., longer processing times per unit). Still, if processing speed isn't a concern, it's a viable solution, assuming your graphic allows you to rotate without creating yet another long horizontal line.

[Dave Sheldrake]

Isn't this one of those times when you could rotate the image so there are no perfectly horizontal lines? This would avoid the situation of the scan direction lining up with your art.Or am I missing something important?

love it!! akin to designing a pen that works underwater when a pencil will do the job just fine.

Sometimes all the physics are academic when the simplest of solutions is the best.


the acceleration happens before it starts firing

In many cases on Chinese machines the acceleration is still happening when the laser is firing, it depends on the speed settings in the engrave section. That's the reason line burning and engraving rarely matches the stopwatch time in LaserCUT's software.

best wishes

Dave

[Mikah Barnett]

Isn't this one of those times when you could rotate the image so there are no perfectly horizontal lines? This would avoid the situation of the scan direction lining up with your art.Or am I missing something important?

Thanks for the suggestion, Dee, but no that's not a really feasible option in this case. I use this for production and typically run 36 pieces at a time. Though I'm sure I could figure out a way to line those up on a diagonal, it is already enough of a challenge getting 36 pieces properly aligned using features available on the part (notice the flat that runs parallel to the horizontal line in the pic - that's my alignment spot). But, thank you for the suggestion!

Seems the power-on time of the laser makes the most sense as the explanation and fits with 100% of what I've observed. I am also going to check the laser alignment as I do have slight difference in thickness between horizontal and non-horizontal lines that I can't otherwise explain.

I also ran a test last night using the Grade Engrave function on a dithered version with 90% black and the problem disappeared completely. I do not yet understand the settings, nor why for Grade Engrave white = laser on while using regular old Engrave white = off, but it is progress. Thanks for all the suggestions - I have already learned a lot, more pictures coming later today...

[Dave Sheldrake]

Polarization of the beam really only has an effect when you get up to very high-power machines (kW level)

Not really Dan, remember Chinese machines are PWM controlled as well rather than PRF controlled like most western machines.

best wishes

Dave

[Dan Hintz]

Seems the power-on time of the laser makes the most sense as the explanation and fits with 100% of what I've observed. I am also going to check the laser alignment as I do have slight difference in thickness between horizontal and non-horizontal lines that I can't otherwise explain.

I also ran a test last night using the Grade Engrave function on a dithered version with 90% black and the problem disappeared completely. I do not yet understand the settings, nor why for Grade Engrave white = laser on while using regular old Engrave white = off, but it is progress. Thanks for all the suggestions - I have already learned a lot, more pictures coming later today...

If the vertical lines are thinner than the horizontal, this is even more evidence the laser is not firing as soon as commanded (i.e., you need more bias). Using a dithered version is continually commanding on/off, giving the laser time to stop firing between dots but not enough off time to back too far away from the lasing ability.


Not really Dan, remember Chinese machines are PWM controlled as well rather than PRF controlled like most western machines.

I'm not sure what you mean, Dave... the firing control method (PWM, PRM, etc.) has no effect on polarization of the output beam.

[Dave Sheldrake]

Hiya Dan,


I'm not sure what you mean, Dave... the firing control method (PWM, PRM, etc.) has no effect on polarization of the output beam.

True but Chinese machines are already at a disadvantage due to them having PWM control (or rather non adjustable ppi) rather than PRF that western users benefit from. Add in the rather "rudimentary" lens's used and the non polarised beam and the quality of work they produce is lower in general. I wouldn't slam Chinese machines (I love the things, they are considered consumable) but the software is a little "less than desirable" in some cases. In general Chinese machines swing round LaserCUT (and it's variations) and gives little in the way of end user tune ability :(

I have heard of different software packages that will work with the leetro control but haven't tried it, however by all accounts some of the packages are reportedly rather good.

best wishes

Dave

[Dan Hintz]

True but Chinese machines are already at a disadvantage due to them having PWM control (or rather non adjustable ppi) rather than PRF that western users benefit from. Add in the rather "rudimentary" lens's used and the non polarised beam and the quality of work they produce is lower in general. I wouldn't slam Chinese machines (I love the things, they are considered consumable) but the software is a little "less than desirable" in some cases. In general Chinese machines swing round LaserCUT (and it's variations) and gives little in the way of end user tune ability :(

Sorry, but I can't agree with you here. The Chinese machines are at no disadvantage based solely on their method of control, other than to say the on/off might not be as finely controlled (though we're not looking at picosecond rise/fall times here, so it's easily ignorable). The optics between Western and Chinese machines are all but identical... they may come out of different factories (for the most part), but there are no special optics installed on our Western machines that aren't also on the Chinese versions. In fact, the Western systems typically used here 25-80W are linearly polarized... it's only when you get into the dual-resonator designs that you start to get away from a simple linear polarization (or someone thinks they need to add in a circular polarizer... which at those low powers you really don't). BTW, you should refer to them as randomly polarized, not non-polarized (no such thing). In some ways, a randomly-polarized beam is beneficial as the cut will be more uniform (though circular is best for uniformity).

As far as the Chinese software sucking, well, yeah, can't argue with that...

[Dee Gallo]

Thanks for the suggestion, Dee, but no that's not a really feasible option in this case. I use this for production and typically run 36 pieces at a time. Though I'm sure I could figure out a way to line those up on a diagonal, it is already enough of a challenge getting 36 pieces properly aligned using features available on the part (notice the flat that runs parallel to the horizontal line in the pic - that's my alignment spot). But, thank you for the suggestion!

Mikah, you don't have to rotate the thing 90°, a small 5 or 10° would take it out of square and still let you use a basically grid like layout.

[Mikah Barnett]

Yep - I understand, Dee - rotating it 90 degrees would actually defeat the purpose (since the now vertical lines would become horizontal!). The problem remains that I need to physically arrange 36 pieces in an array that is square with the same lines the logo is aligned with.


Mikah, you don't have to rotate the thing 90°, a small 5 or 10° would take it out of square and still let you use a basically grid like layout.

[Mikah Barnett]

Kinda wanted to put a period on this - forum posts with problems stated and no solutions are a pet peeve of mine (assuming a solution is found).

After reading the comments here, I decided to re-check all my hardware and I found a couple big issues:
- Mirror alignment was good to the final mirror, but the final mirror was not shooting at the center of the lens and had no adjustment left
- My 63.5mm lens was toast - cutting Polyurethane Foam with no blower on was a bad choice and cost me a lens

So, I removed the final mirror and added some spacing material to give me some range of adjustment, then I adjusted it to shoot through the center of the lens - it was off by about 3mm to start. I also replaced the lens with my 50mm new lens since I had ruined the other. In the process, I discovered that I could change the head height (distance from final mirror to lens) on my machine which allows me enough clearance to cut 1.5" thick PU foam WITH the blower nozzle in place - that should help avoid future issues.

So, I fixed those issues, fired it all up, and had the same problem. Bummer. I did a ton of experimenting with LaserCut 5.3 Grade Engrave functionality - enough to develop a pretty thorough understanding of the ins and outs of it and I was able to cause some different behavior, but nothing really right.

In the end, I finally found one solution that seems to have cured 90% of the problem: FOCUS!!! If the beam is just a little out of focus - and I mean fractions of a mm - I get significantly more etch on longer horizontal passes and less on short ones. If I get the focus just right on the money, it's nearly perfectly consistent regardless of the shape I'm etching. Still very sensitive to power level - a couple of % one way or the other on my 80W tube and I vary between not fully etching and etching right through the anodic coating. When the etch does go through, it starts first on the horizontal lines still.

Anyway, for now I'm able to make it work repeatably and properly as long as I take extra time to get the focus exactly right. Thanks for all the suggestions here - it has been an education!

[Dan Hintz]

Glad you got it (mostly) figured out, Mikah...