Using fiber laser engraver attatched to an existing cnc machine ?

[Rab Gordon]

I've been looking at fiber laser engravers as they seem to be able to engrave finer detail than I can achieve with a high speed spindle and they have pretty reasonable prices for say a 30W machine.
However, the software that typically comes with them seems to cause users a lot of issues for rotary engraving, never mind full 4 or 5 axis work.
So, I'm wondering how feasible it would be to attach a fiber laser source with simple fixed focus optics onto one of my existing cnc machines similar to how many users have done using a diode laser ?
Are there any particular issues with doing this or in controlling a fiber laser source ?

thanks,
Rab

[Nick Cicala]

I've been looking at fiber laser engravers as they seem to be able to engrave finer detail than I can achieve with a high speed spindle and they have pretty reasonable prices for say a 30W machine.
However, the software that typically comes with them seems to cause users a lot of issues for rotary engraving, never mind full 4 or 5 axis work.
So, I'm wondering how feasible it would be to attach a fiber laser source with simple fixed focus optics onto one of my existing cnc machines similar to how many users have done using a diode laser ?
Are there any particular issues with doing this or in controlling a fiber laser source ?

thanks,
Rab


you'd have some complex problems. Focusing would be an issue. Are you processing metal? Fiber only affect non-organic material. What kind of application would you be doing/ what purpose of putting this on a cnc machine, also how big of a CNC machine?

[Rab Gordon]

Primarily I'd like to make rings in titanium, silver, gold, etc.
I currently do this using a high speed spindle on a small milling machine.
I'm assuming that focus could be set by adjusting the height of the lens above the workpiece, attached to the Z-axis.

[Nick Cicala]

Primarily I'd like to make rings in titanium, silver, gold, etc.
I currently do this using a high speed spindle on a small milling machine.
I'm assuming that focus could be set by adjusting the height of the lens above the workpiece, attached to the Z-axis.

Personally I think you'd be better off with a stand alone system, adding a laser into a CNC area is no bueno, you can run the risk of damaging the optics with debris. I'm sure you could find a z-post 20w stand alone unit instead trying to rig something up.

[Rab Gordon]

Would there be any more danger to the optics than a typical scanhead fiber laser ? Would it have to be a much shorter focal distance ? If so then could you add a cutting head style shroud with air purge to keep the lens clean ?

I found one DPSS metal engraving laser designed for this type of application, although only 10W and a relatively large spot size;
http://endurancelasers.com/an-endurance-brand-new-fiber-marking-module-for-metal-engraving/

(http://endurancelasers.com/an-endurance-brand-new-fiber-marking-module-for-metal-engraving/)Was really hoping to find a similar 30W fiber laser source.

Rab
cnc-toolkit

[Rab Gordon]

Should I have posted this in the Engraving-Hardware-and-Software sub forum ?
Can it be moved ?

Rab

[Kev Williams]

I'm not sure exactly what your intentions are--
If engraving, as long as your parts are sized below the working area, which is 'adjustable' via lens size, you don't need a gantry...
If machining, which is very tough to do with a fiber, you can't use a galvo head laser, since the beam- aka your 'cutting tool', is always on an angle, no straight cuts. You'd be better off looking at a gantry-steered fiber, like the Epilog. They run at glacial speed compared to a galvo, but it's the only way I can think of to make a straight cut...

I believe Epilog (for one) also makes a gantry-galvo combo machine with a 24x24" work area, but it's spendy- I'd put up a video but can't find a decent one (why a company selling a 24x24 machine will only show 4x4" of it working is beyond me?)

I've been thinking of fabbing up a bracket so one of my old 5000XT's (rotary tool machine) would hold the scan head & housing to the spindle housing, so that I can fiber etch the 22 x 31" operator panels I make; I'd just use the XT to move the scanhead around manually. It would work, and I could focus easy enough since these spindle heads have about 3" of adjustment vertically. Just haven't had time to actually do it. And FWIW, the last thing I'd worry about is messing up the optics ;)

Something like this moving XY table would help me, but I'd need one that would move at least 30x30", and to do that would require fabbing up a boom with 30" of clearance from the lens to the tower. But that's do-able. Would something like this work for you maybe?


https://youtu.be/GmkKETpiQBY

[Rab Gordon]

Hi Kev,

Thank you for your thoughts on this.
I'm primarily wanting to engrave rings, however I see a lot of posts of people having various issues when using a rotary axis on their fiber engravers.
Seemless engraving often comes up as causing problems and engraving by following vectors seems to be surprisingly difficult.
Which is why I though of incorporating the laser with one of my current cnc machines which cope quite easily with rotary jobs such as rings and can also manage more complex curved surfaces using full 4 or 5 axis mode if needed.

One thought would be to get a typical off the shelf, galvo head fiber engraving machine and start simple;
Use ezcad to set power and frequency and fire the laser at a single point, but find a way to interrupt the laser fire signal to turn the beam on and off externally, ie. treating it like a spindle from the cnc machine.
I doubt it'd be as simple as that, but good enough in theory.

For your big panels, I saw mention in a bjjcz manual of repeating the last job held in memory, so if you're engraving the same thing at different locations on your panel that would seem possible to automate by moving to location then using an output from the 5000XT to trigger it's engraving cycle ?

Rab
cnc-toolkit

[Kev Williams]

Rings are easy to engrave with a basic rotary and EzCad, as long as it's just text. For inside engraving hold with the jaws, then move the rotary to the left so the ring is about an inch from the lens center. Draw a small box a bit wider than the ring, then redlight it, and position it so it hits the bottom inside of the ring. Experiment with how far you have to move the box and/or the rotary left from center so that the beam will fully hit the engraving area at bottom of the ring but misses the top of the ring. This is where the beam angle of a galvo HELPS :) EzCad always home-positions the rotary from vertical dead-center of the work area, as such, center your engraving vertically.
There's a trick to remember when engraving rings, a little thing called 'INVERT'-- this is necessary because the machine will engrave text in reverse order on the inside of rings if this setting isn't correct, same with engraving the outside, nothing sucks worse than seeing "!uoY evoL I" engraved in someone elses ring ;)

Engraving the outside, I use a tapered mandrel and hope ;) - but I don't do many rings. Again, text is easy using the 'rotate text mark' function (I don't use the ringmark function), the rotate text function automatically moves the rotary for each letter, no fuss. If you intend to engrave graphics on rings, that can be problematic if the graphic is longer than can be engraved without moving the ring-- and a CNC won't help with this..

I've done a few of these, I've gotten better over time, this was the first set of rings I ever lasered. As they say, if I can do it...
416612

To address the 'repeat job' in EzCad, (they call it REMARK) I discovered this function accidentally when getting my first Ebay laser. In your example of engraving my panels, the machine will repeat 'normally', just press the F2 button again! The purpose of the repeat job, best I can tell, is this: When engraving multiple parts, every time you start the machine, you disable EzCad, it is now busy streaming your engraving data to the controller, and awaitng a STOP command from you in case you need to stop. Once the part is done, EzCad is usable again. In many cases it's not a big deal, but suppose you're engraving 30 parts, and they take 30 seconds to engrave, and you have another job to setup; you can't setup the job while running the first batch of parts. But if you're using the 'remark' function, the machine engraves from the controller memory instead of your computer, so now your EzCad is sitting idle, and you can now save the job you're working on and set up your next job! Depending on the work you do, this can save hours a day in setup time! To make the remark work involves wiring up a separate start button or foot switch to the control board, piece of cake if you know how to solder. There is ONE catch to the remark function, you CAN'T STOP the engraving, other than by using the Estop switch. However, that's rarely an issue because using the remark assumes you have everything right as far as the engraving goes. But that doesn't help with putting parts on the machine upside down! ;) That all said, I've never needed the remark function, since I have 3 fibers I'm not usually waiting on the one that's running...

[Rab Gordon]

I can see how text would be relatively easy to engrave on a ring, however I engrave continuous designs which would have to be seamless throughout and to meet up seamlessly at the start/end of the design. If the ring moved around and the laser beam remained stationary that would bypass any limitations in ezcad's rotary capabilities.

Sorry, but for some reason I can't view your rings photo.

Rab
cnc-toolkit

[John Lifer]

To view Pics, you need to be a contributing member, $6 per year min.
I'll piggy back onto Kev, I've done probably a dozen rings. And probably that many more round objects.
Text is easy, most other stuff is difficult. I've added small vector images to several rings, usually either side of the text.
I've got local jeweler that I've worked with, She is a great engraver with decades of experience, but usually only does text now with drag.
So I've got simple designs that I've helped her with.

For your continuous engraving, good luck. I would say if you only have a few designs and can make each a separate file for each ring size, it will work well.
If you are doing custom one off designs that change, it is more of an issue.
I think stand alone machine is your only choice, and a 20 watt is fine for rings. I can cut plenty deep in silver and gold and titanium.
Here are a couple I've done.416691
416690

[Rab Gordon]

Thank you John,

Still new here, I've just set up the annual subscription and can now see Kev's and your photos, they're nice clean engravings.

Here's a couple of the rings I've made, all custom commissions;
416692

416694

416693

Video; https://vimeo.com/361806033
(https://vimeo.com/361806033)
Rab
cnc-toolkit

[Kev Williams]

The issue with rotary engraving 'long' graphics onto a cylinder shaped object using a galvo steered laser is the seams between 'splits'- because no matter how hard you try, it is virtually impossible to perfectly align the beam to the object- 1/100th of a degree of skew in the shape of the object itself or the angle of the rotary to the machine, or 1/100th of a mm of out-of-round, or 1/100th of a mm deviation in diameter, will result in seam misalignment--
416696
This is an SS water bottle I was experimenting with- note the seam lines, and to my point, you'll noticed the seams change size- in the bottom engraving toward the left side alignment was very good, but not so good on the right, the top engraving is horrible, and the seam spacing grows the farther left of the image, which is the opposite of the bottom.

Now then, with something as small as a ring, this anomaly would be much less pronounced...

There is a band-aid fix some use to counter this problem, which is to make the split distance very short, like .1 or .05mm instead of the 2-ish mm split in my example. This is a way to 'artificially' mimic the Y-axis movements of a gantry. It works, but because the laser has to wait for the rotary to move it drastically increases the engraving time. However, in your case, where you're likely not mass producing parts that are always waiting on the machine, who cares how long the engraving takes? The beauty of a laser beam, you never have to stop to sharpen it! :D

I want to add to the use of the 'rotate textmark' function some may not be aware of-- you don't need text for it to work :)... It will auto rotate and engrave ANY single object. Note that there's NO 'split' options in 'textmark', all rotation is automatic... example, suppose you draw up a ♠♣♥♦ to be engraved around a cylinder, IF you hatch fill all of them as a group, 'textmark' will engrave them all at once with no rotation. BUT, if you hatch each one separately and run it, the laser considers each the separate entity that it is, and will rotate to engrave each one separately, automatically. And the laser doesn't care where each entity is located, which is not the case when using 'splits'.

Suppose you wanted to engrave these scrolls around a ring, note all scrolls are separate-
416699
using normal rotary modes, there's no way to avoid cutting the engraving in half because the scrolls overlap each other in the horizontal plane - But textmark will engrave each scroll separately, no seams :) ...








.

[Rab Gordon]

Thank you Kev, that's very useful information.

I did find a youtube video of rotary vector engraving which would be seamless, but haven't been able to find out any more about the "BJJCZ marking while rotating board" used;
https://youtu.be/CjTYmt3ExbE

Rab
cnc-toolkit

[Kev Williams]

cool-- I see one caveat, you won't be able to do that while rotating a 4 pound or so chuck! But there's ways around that of course ;)

and the 'must use BJJCZ marking while rotating board' means (to me) that a different controller board is necessary. But will this board also function as a normal board? Hmm..

>edit<

Found this- note that they recommend a servo motor. Also note, price not listed ;)
416714


.

[Rab Gordon]

It does seem to be an add on board, but no doubt not cheap and with it's own problems.....
Strangely difficult to find more info or a manual for it.

I would have thought that such functionality would have been integral to any standard control card which has rotary functionality.

Rab
cnc-tookit

[Tony Lenkic]

Take a look at this site.........EWC Group Inc.
These people do rings you are are doing and they also sell lasers to do it.
Check them out.

[Rab Gordon]

Take a look at this site.........EWC Group Inc.
These people do rings you are are doing and they also sell lasers to do it.
Check them out.

Thank you Tony, I wasn't aware of them, I'll take a look at their site.
Rab

[Nick Cicala]

Hi Kev,

Thank you for your thoughts on this.
I'm primarily wanting to engrave rings, however I see a lot of posts of people having various issues when using a rotary axis on their fiber engravers.
Seemless engraving often comes up as causing problems and engraving by following vectors seems to be surprisingly difficult.
Which is why I though of incorporating the laser with one of my current cnc machines which cope quite easily with rotary jobs such as rings and can also manage more complex curved surfaces using full 4 or 5 axis mode if needed.

One thought would be to get a typical off the shelf, galvo head fiber engraving machine and start simple;
Use ezcad to set power and frequency and fire the laser at a single point, but find a way to interrupt the laser fire signal to turn the beam on and off externally, ie. treating it like a spindle from the cnc machine.
I doubt it'd be as simple as that, but good enough in theory.

For your big panels, I saw mention in a bjjcz manual of repeating the last job held in memory, so if you're engraving the same thing at different locations on your panel that would seem possible to automate by moving to location then using an output from the 5000XT to trigger it's engraving cycle ?

Rab
cnc-toolkit



This can be difficult depending on the software. I do this frequently on a Trotec Speedmarker 1300 30-watt galvo-fiber system. The software has a segmentation feature which enables rotary engraving pretty easily. Seamless engraving accuracy depends on the the correct diameter value put in the software. If its nominal the graphics can be off but it doesn't take much to recorrect. It all depnds on how serious are you? The SM-1300 is an amazing workstation and is built for high-volume industrial marking.

[Rab Gordon]

This can be difficult depending on the software. I do this frequently on a Trotec Speedmarker 1300 30-watt galvo-fiber system. The software has a segmentation feature which enables rotary engraving pretty easily. Seamless engraving accuracy depends on the the correct diameter value put in the software. If its nominal the graphics can be off but it doesn't take much to recorrect. It all depnds on how serious are you? The SM-1300 is an amazing workstation and is built for high-volume industrial marking.

Hi Nick,
I'll check it out.
Thank you,
Rab

[Nick Cicala]

Hi Nick,
I'll check it out.
Thank you,
Rab


Here's a video of a testing I just did, it was for a full marking around a medical instrument. It's small, but you see the segmentation and how it works. It you notice some of the other marking that are blocking and not constant, that is a result of the wrong diameter being put in.

https://www.dropbox.com/s/h0mq85jpsnr5zv9/MVI_0904.MP4?dl=0 (https://www.dropbox.com/s/h0mq85jpsnr5zv9/MVI_0904.MP4?dl=0)

[Rab Gordon]

Here's a video of a testing I just did, it was for a full marking around a medical instrument. It's small, but you see the segmentation and how it works. It you notice some of the other marking that are blocking and not constant, that is a result of the wrong diameter being put in.

https://www.dropbox.com/s/h0mq85jpsnr5zv9/MVI_0904.MP4?dl=0 (https://www.dropbox.com/s/h0mq85jpsnr5zv9/MVI_0904.MP4?dl=0)

Thanks,
How would this segmentation method work when engraving a ring with a curved (D-Shaped) cross-section ? The diameter is larger in the middle of the ring and decreases as it goes towards the edges.

Rab
cnc-toolkit

[Kev Williams]

That will be problematic, you would just about HAVE to run very short spaced "micro-splits", and even then actual engraved hatch lines will vary in spacing from programmed spacing. But done right, it could work, but it would be slow going...

And I just wanted to point out, the segmentation thing shown in Nick's video is pretty much identical to how the Chinese machines do it, it's a what I referred to as "splits" in my last post, each split represents a different segment, etc...

[Rab Gordon]

That will be problematic, you would just about HAVE to run very short spaced "micro-splits", and even then actual engraved hatch lines will vary in spacing from programmed spacing. But done right, it could work, but it would be slow going...

And I just wanted to point out, the segmentation thing shown in Nick's video is pretty much identical to how the Chinese machines do it, it's a what I referred to as "splits" in my last post, each split represents a different segment, etc...

Is there a reason for it being common to split / segment the design rather than following a vector as would be done with mechanical engraving and which you'd think should work better for this type of engraving ?

Rab
cnc-toolkit

[Kev Williams]

Below is the video from the link on the previous page- That setup will do the -on-the-fly- engraving you're after. Probably a $1000 or so upgrade, maybe a bit more because to get the best results you'll also need to replace the stepping motor with a servo motor.

The reason the new setup will work while the 'basic' rotary won't, is because (a) the basic rotary's controller isn't designed the 'steer' the rotary on the fly, and (b) I'm thinking the basic steppers may not be capable of the microstepping necessary for high-quality work. That's the main reason for segmenting the engraving...

The video's comments mention a servo motor should be used... And regardless, note in the video there's no heavy steel chuck, just the motor and the part. It would be a pretty tough task to rotate a chuck at the speeds a fiber is capable of running.

I think this video is really cool, I'm giving serious consideration to getting one of these, and possibly a 3D setup for one of my machines :)


https://youtu.be/CjTYmt3ExbE

[Rab Gordon]

Below is the video from the link on the previous page- That setup will do the -on-the-fly- engraving you're after. Probably a $1000 or so upgrade, maybe a bit more because to get the best results you'll also need to replace the stepping motor with a servo motor.

The reason the new setup will work while the 'basic' rotary won't, is because (a) the basic rotary's controller isn't designed the 'steer' the rotary on the fly, and (b) I'm thinking the basic steppers may not be capable of the microstepping necessary for high-quality work. That's the main reason for segmenting the engraving...

The video's comments mention a servo motor should be used... And regardless, note in the video there's no heavy steel chuck, just the motor and the part. It would be a pretty tough task to rotate a chuck at the speeds a fiber is capable of running.

I think this video is really cool, I'm giving serious consideration to getting one of these, and possibly a 3D setup for one of my machines :)


I think this upgrade would be a big improvement, however I've not been able to find out much more about it or find it's model number or read the manual.
I've been told that the controller from scaps is far better but it's also far more expensive....

Rab
cnc-toolkit

[Nick Cicala]

Thanks,
How would this segmentation method work when engraving a ring with a curved (D-Shaped) cross-section ? The diameter is larger in the middle of the ring and decreases as it goes towards the edges.

Rab
cnc-toolkit


Depends on the lense used, I stick to a 160mm for testing and if I need to go bigger because due to marking size or focal range it's 254mm then 330mm but the beam diameter increases heavily and results might not work depending on the accuracy needed. The segmentation or splitting is a issue when it's a full wrap graphic without any breaks, you can manually set each segment as well to dodge splitting a letter or graphic in half. I wish non segmentation rotary would work but it just can't for engraving and single point outline has limited applications.

[Kev Williams]

So while I've mentioned 'micro-splits' above, I've never actually tried it-- till now :)

Just finished a job with the rotary and decided to experiment. Here's what I did-

First, I wanted to have the rotary microstep to its nearest equal 'built in' stepping distance, so- I changed my rotary's step driver to rotate 20,000 steps a long time ago. The part I engraved is 1.25" diameter, which is 31.75mm, times pi = 99.7458mm, divided by 20,000 steps = .00498729mm per step. Ok, so .005mm should work. HOWEVER- as I type this I just realized I'd factored the stepping using the 31.75 diameter rather than the 99.75mm circumference-! This gave me a .004mm distance... So regardless of that, the test came out okay, but I'm wondering if I'd used .005mm as a reference if this would've turned out better...

anyway--

So using the .004mm reference, to start I used a .08mm 45-135 degrees cross hatch, and I used a .16mm split size; the hatch and split sizes being multiples of .004--
I then engraved the top 'Liberty' graphic...

For the bottom graphic I used identical settings EXCEPT that I changed the split size to a full 2mm...

Results:
417086
Not too bad really, as far as normal viewing goes. But note the top version is not as white as the bottom version, there's a reason for that...

Closeup of the top version, notice the vertical lines, they appear to be hatch lines but,
they're actually all the .16mm spaced split lines...
417087
But the engraving and image in general looks very nice, but the hatch engraving is
hidden BY the split lines, which is why this engraving doesn't appear as white, the hatch engraving
reflects much more light that the residual split lines...

Closeup of the bottom version, the 2mm split lines, while faint, are clearly visible. The only
benefit to this engraving is it's mostly hatch fill, and brighter. But the split lines are the deal breaker...
417088

Took about 40 seconds to run the top version, and about 10 seconds to run the bottom, give or take...
========================

And as I think about the 'vector rotary' above, I'm not real sure it'll work well with cross-hatch engraving, because, no matter how fast or accurate it is, it's taking the place of a mirror capable of moving engraving 10,000 mm/second. Even if the rotary is capable of 5 revs per second, at 1" diameter that equates to 400 mm/second, 96% less than galvo speed... So I may rethink the need for one ;)

[Rab Gordon]

Is the speed mainly an issue due to the increase in time the job will take or it is more of an issue that the beam needs to be moving fast to get a better quality engraving ?

Rab
cnc-toolkit

[Kev Williams]

Speed of engraving directly correlates with engraving quality; a lot of fast passes to the same depth will *usually* result in better engraving quality than a few slow passes. However, for the most part, the quality (IMO) refers to the EDGES of the engraving. Engraving within the edges usually isn't a problem, and if it is some cleanup passes at different settings can take care of it. But the edges are a different story, when cutting out the middle the hatch overlap pretty much vaporizes/removes the metal, but the edges are the beam's stop point, and the hot metal builds up in the form of slag. I liken fiber laser engraving to cutting steel with an acetylene cutting torch, just on a MUCH smaller scale---
417113
I just grabbed this pic online of cutting torch tests, a bit of an oversimplification but it does describe what's happens during lasering with a fiber: the middle goes away but part of the middle remains on the edges.

The trick is to come up with an engraving routine that will hog out the middle while tempering the engraving along the contour edges...

The reason my micro-split engraving took 4x as long to run is simply because the laser had to stop and wait for the rotary to turn. And while my engraving of the top version of my test looks good, the sad fact is, the vertical lines are all just individual stop points of the laser beam, aka 'new edges'... so all those vertical lines are actually slag buildup. With a lot of testing with different engraving routines, I could probably greatly reduce the slag, but how much..?

[Rab Gordon]

Thank you Kev, a good explanation.

Rab
cnc-toolkit

[Rab Gordon]

Would it be possible to slow the pulse frequency right down, keeping the power per pulse the same, such that the number of pulses per inch was the same for the beam traveling slowly (ie. doing the engraving while rotating method) as for the beam traveling very fast using just the galvo head ?
If so then would the results be much the same ?

Rab
cnc-toolkit