Hopefully no one has said this already as I didn’t read all of the responses. My experience with cutting wood is that if you don’t get all the way through it on the first pass, you have created charcoal in the curf, which is exponentially more difficult to cut through. I am not 100% on this but logic says that a wider beam would not give you as much energy concentrated in the cut diameter than the narrower beam. Since you aren’t going through thick material, the smaller beam (shorter focal length) would be the better choice in my opinion. That being said, some wood species will be more forgiving on multiple passes but you probably want to get all most all the way through on the first. I cut a lot of plywood, which can cut nicely depending on the grade but with my 100 watt, about 3/8" is the most I can get through. Baltic Birch ply, forget about anything thicker than 1/4" and even then sometimes it doesn’t go all the way through even with multiple passes. The problem becomes when you try to slow down your cut too much. In theory, you should be able to get all the way through thicker material with a slower cut but what usually happens when I try it is that the wood starts smoldering inside the cut and burns out a hollow space in the material, ruining the part. It has been a while but I think I remember trying wetting the piece to mitigate that but it didn’t help. The only thing that helps the cut seems to be air pressure. I usually run somewhere between 20 and 60 psi when cutting wood. FYI, just the opposite with acrylic. Set your air flow to the absolute minimum only to keep the lense clean. Keep a spray water bottle handy as it likes to flame up and sometimes won’t self-extinguish by pausing the cut. Hope this helps
There is a lot of contrary opinion on this. I have cut 10mm bamboo in 10 passes with a 10w diode laser. No way I will try it in one pass.
Hi Mike,
I am not familiar with diode lasers so I can’t comment on them. I have had this experience on my 100w co2 laser. Bamboo cuts and engraves much better than Baltic Birch plywood. I think it has to do with the glue that is used and the density of the material. Bamboo is technically a grass and (I am assuming you are talking about Bamboo ply. ,“Plyboo”, etc.) the glue they use is a natural glue. The darker colors are also baked, which is how they get the darker colors all the way through the sheet, so that may affect how dense the material is. Someone did comment that the diode lasers don’t create the same problem as the co2 lasers. I’m guessing that it is because it doesn’t get as hot?
My single diode laser 5.5 watts can’t cut nicely enough through relatively thin material. Several turns are always used with charred edges as a result. But it is primarily because the laser is too weak and not suitable for these kinds of tasks. (and believe me it gets hot)
Regarding glue, if no laser suitable glue is used in the wood, you cannot cut nicely or not cut through plywood at all. This applies to all diode and the CO2 laser I know, no matter the effect of the machine.
I have some 6 mm birch plywood that just can’t be cut through without charred edges. It does not matter if I use a single or several turns, it actually gets worse. When I have to cut this material anyway, then all the compressed air I can produce must be used and the result is still not good.
Expecting a laser to cut without char is like expecting a bandsaw to cut without making sawdust.
I had a 10mm thick bamboo cutting board that I cut with 10 passes and a 10w diode, hit or miss. I took it to a local laser shop that tried to cut it with a 30w CO2 laser. I needed the outline of a shape to be cut. We quit after 20 passes did not cut through the board.
There seems to be no “standard” for what works and what does not. I cut the board with a 10w laser, the CO2 did not, so I chose to go for a more powerful diode machine. It does the job without having burn-out spots caused by using low speed… Go figure!
It is possible, but there are boundaries that I want/can’t accept in my work.
The meaning of my post was that there are materials and or deliveries that are not worth cutting/ processing with - my laser (diode and CO2). Everything becomes Dung Dirty and post -processing takes way too long.
The picture shows my limits and the only difference is glue. If the newer diodelasers can handle the task with exterior glue better then, that’s fine with me.
In the bottom picture there are fire marks on the cuts parts, but if the color or soot does not rub off than that’s fine too.
Here it is not glue or other chemicals that give discoloration, but lack of laser powers for the task, it was only a test.
I will have to put a test for this on my ToDo List.
…I wrote “if”…
Can’t determine “if” without testing, right? Besides, I like a good challenge, and sometimes even bad ones. 
…just so I’m in and don’t lose anything in translation, what are you going to test? 6mm plywood with external glue?, I’d love to see a test on that, made with a diode laser of the newer date, go for it and show the result. 
Nothing lost. I have some 1/2" marine it home I can test on with the SF-A9. About 3 weeks out because I am with family in NY now.
Of course I normally use a 2" lens. But I have a 4" lens and have done some testing with that. I test things A LOT just to know what is possible.
In almost all cases, the 2" cuts better. I cut all the way through a spruce 2x4 with a nice clean cut in one pass. Walls pretty vertical and the kerf is small enough that the part may not fall out freely.
This is the paradox- the beam should diverge past the focal point, right? If I have the fixed at the right height above the bed to focus the cut on the top of 1/2" plywood, it cuts. Hey, wait- if I take out the 1/2" plywood and put in 1/4", and don’t move the focal point, why can’t I cut the 1/4" at all? It just makes a wide burn mark. After all, isn’t 1/2" the same as two 1/4" sheets glued on top of each other? What magic is that top 1/4" doing that lets the bottom 1/4" cut?
And here’s the answer, that most people have a hard time accepting- 10.6um CO2 lasers REFLECT off the cut walls at very low angles of incidence, and it’s surprisingly low loss on each reflection. Once a basic channel forms with near-vertical walls at the top of the material, then the diverging beam reflects back and forth off one wall and the other until it hits the bottom of the channel, which further deepens the channel and hopefully still makes clean vertical walls so this keeps going on and on.
The channel that forms has a “mostly” fixed width for its entire depth. Actually, it’s not entirely constant, it’s actually got some really interesting geometry in how it changes with depth and forms its reflection funnel but to the user that variation is small and not really important to making a product.
The initial channel form has to happen within the focal waist of the beam to create the parallel vertical walls, and this geometry will “capture” the beam into the channel. If the channel does not form in the required shape, it will NEVER form and beam capture will not happen. That is, you start out of focus and make a big burn mark, it will never cut. If you bandsaw it to see the cross-section, it’s a wide crater that cannot funnel the light into a narrow channel. Giving it higher power or going slower will never create the necessary funnel geometry of near-vertical walls that channel the beam. You will never get a cut.
This may bring up recollections of fiber optics and TIR “Total Internal Reflection”. But TIR is a phenomenon of trying to escape from a high index of refraction material (glass) to lower (air or a lower index type of glass). That’s actually the opposite of our case- the air is low index, the material (even carbon) does have refractive index and it’s higher, not lower. Rather, this is essentially glare- and both occur at very low angles of incidence, so the difference is kind of pedantic. It’s like the glare off water where you’re looking off in the distance- for light which is nearly parallel to the water, the reflection is near total.
But, a 4" lens means the beam’s divergence angle past the focal point is half, so it should be EASIER to capture, right? Well, it doesn’t. I do think there’s more to it than I understand well enough to spell out yet, but one basic problem is lenses have to follow a “diffraction rule”
The focal point is never actually a zero-width point anywhere. It’s a focal waist with a minimim diameter aka “spot size” that follows the diffraction equation:
2w0 =(4M^2λf)/(π*D)
f is focal length. There are some numbers here which won’t matter- λ is wavelength which is fixed at 10.6um for CO2. M is a mode parameter of the laser source which is near 1 and doesn’t vary by a lot from laser to laser so don’t worry about that. With a 2" lens, you get about 0.125mm of spot size. If you go with a 4" lens, the spot size doubles in diameter, which means the energy density decreases by a factor of 4. This generally makes it harder to ablate material into the funnel shape and start a channel. Also, the depth of focus on 2" is +/- 2.5mm, but that’s highly subjective how far being out of focus is acceptable. On 4", the tolerance is +/-5mm.
Incidentally, this is why some people use a 1.5" or even 1" lens for fine engraving. The diffraction rule means the spot size gets smaller, so it can do finer lines, and also the energy density increases which may be necessary for lower power machines to be able to engrave the anodization off aluminum.
Anyhow, bottom line, I tried with wood and the cut was always slower, wider kerf, and the ultimate depth I could get through at any speed was less with the 4". I don’t think I’ve done enough with thick acrylic to say the same, I need to try that,
The ONE material case where the 4" shines was 1.5" 2# EVA foam. Actually, I could get through it with the 2" too, but after about 0.75"-1" the cut has to be slowed down disproportionately slower to get more depth. It was a crawl to get to 1.5". The walls weren’t really vertical and it was getting pretty melt-y. In this material, with a 200W source, I got 27mm/s and really clean, vertical walls on the 4".
Interesting note for those with a curious mind- that diffraction equation has something “crazy” hidden in it. If you start with a beam of half the diameter when it enters the lens, you’d expect the spot size to be half, right? Nope, this is one of those things that will mess with your head. INCREASING the beam diameter- doubling it- is what halves the focal point spot size. The focal point will still be 2" from the lens regardless.
Furthermore, if you think about it, well, isn’t that whole channeling thing about the angle of the light cone as it converges to the focal point and then diverges at the same angle in free air? Yes. Yes it is.
But then, if I double the beam diameter going into the lens, what diameter is it as it exits the lens? For all practical purposes, it’s the same diameter on both faces of the lens. And that means things.
OK, but… wait… would that mean that if I put a 6mm beam going into a 2" lens, if I had a 3mm dia beam, the spot size doubles, but the angle of the focal cone is halved… isn’t that essentially the same cutting dynamics of a 4" lens? The work still has to be focused at 2" from the lens on it, but it sounds like the focal cone is the same, the spot size is the same, and the depth of focus- how it performs when you’re at +/- 1mm from the ideal focal point, and how far you can be “off” due to warped plywood before it won’t cut- sounds exactly the same. Well, it IS exactly the same!
The input beam size matters, just as much as the lens focal length. A beam expander is how you increase beam diameter. Unfortunately, the reverse generally doesn’t work- trying to compress a larger beam into a smaller beam ends up with a divergence problem that really isn’t going to work on a gantry laser system where there’s a significant distance to the next lens.
In the cheap, common glass HVDC-excited tube machines, you rarely see beam expanders. Much to my chagrin, none of these tube mfg- Reci, SPT, etc, have ANY real data sheet for any of their tube sizes. You won’t find a figure for beam diameter anywhere. I can say it’s roughly 1/4" (6mm) and seems to be the same across all the sizes.
Universal Laser Systems’ “HPDFO” engraving optics does that, though. Well, a full beam expander is a negative-focal-length lens to diverge the beam, followed by a regular positive lens to straighten out the beam at a larger diameter. ULS’ HPDFO just starts with a negative spreader to get the beam diameter 2x or 3x by the time it reaches the actual focal lens. In this case, the actual focal point of the beam is 2" below the lens, but the lens type is actually shorter than 2" because it has to both counter the divergence from the prior lens in addition to the adding normal convergence.
Bottom line is HPDFO creates a fine engraving beam even smaller than a 1" lens- it’s actually the equivalent of a 0.4 inch focal length! But physically it’s doesn’t have to be so dangerously close to the lens, it’s at 2". BUT, it still has ALL the liabilities of a short engraving lens- the depth of focus tolerance maxes out at 0.7mm, it has to be very flat and well-focused, and the focal cone angle converges and diverges sharply so it’s not going to form a channel and thus really has no ability to cut.
So, coming around to the bigger picture for conclusion- focal spot size, depth of focus, and the ability to form a channel are inherently linked together by the diffraction equation. A wider input beam does inherently the same thing as a longer lens, they’re inseparable by the laws of physics.
The other takeaway is hidden in that diffraction equation- wavelength. Most CO2 is 10.6um. Well, except, there are RF-CO2 lasers made to fire at 9.6um or 9.3um. These actually do focus to a smaller point, or, with a longer lens, have a narrower focal cone with the same spot size and thus should be able to cut better. And it means when you’re looking at a fiber laser with a ~1um wavelength, or a blue laser diode or UV laser, the focal situation is completely different.
Aw, well, let’s go all in and finish this- the other benefit of longer lens is that “tolerance” figure.
Let’s look at two cases:
- A 6mm dia input beam, 2" lens. That calcs to 0.114mm dia spot size. The thing to note is that the focal cone is 3mm radius at the lens, and essentially zero at the focal point. So the focal cone converges and diverges at tan-1(3/50.8)=3.379deg half-angle. It’s going to be simpler to think in terms of half angle and radius, and that spot size is 0.057mm radius.
- A 6mm dia input beam on a 4" lens. 0.229mm dia spot size but a 1.69deg half angle. The spot size radius is 0.114mm
Say I have a curved surface, but no fancy auto-focus system the adjusts the Z up and down. This is also the case with a rotary working on a curved drinking glass. I can reangle the rotary for best level, but this glass isn’t a cone shape like a martini glass- a curve is still a curve. The distance to the lens will vary.
The actual beam radius as you get away from the focal point is (spot size+(mm of error)*tan(beam half angle).
So when your surface is +/- 2mm from ideal, the 2" lens beam is 0.175mm radius at the surface, whereas the 4" lens is 0.173mm radius.
That’s the crossover point- once you are more than 2mm from the focal point, the beam dia and energy density of the 4" gets better than the 2".
And, ok, let’s just say I’m sloppy. I can’t get my plywood totally flat. Well, if you’ve got more than 2mm of warpage, you really need to rethink the way you hold stuff down rather than going with a 4" lens. On the other hand, there are 2.5" and 3" lens options, you could try to make the case that trading off the smaller perfect focal point min focal spot size for less increase as you move away from the focal point could help your case. That’s a matter of how you use your laser.
Most people still get the best performance out of 2" on flat work, though. Curved glasses on rotary, or, like, trying to engrave the top of a cow skull where you will just have to live with a curve are special cases
Now THAT was some very interesting reading! Thanks for taling the time to write it up.
Finish up with the part for us diode people?
Those are much shorter wavelengths, the blue diode laser engravers have a 23.6x ratio with CO2. This means the ratio of min spot size vs depth of focus for that laser is 23.6x higher.
The diode could get the same 0.125mm spot size of a CO2 on a 2" lens with a depth of focus of +/-59mm. So in theory it could engrave all over a cow skull without going out of focus, with that spot size.
They don’t actually make them like that, though. I don’t have specifics on it, but they’re a much smaller spot size than CO2. It needs to be, because they’re so much smaller wattage than CO2 and energy delivery is much weaker. The depth of focus is similar to CO2, the performance will be trash if it’s off by a few mm.
By “energy delivery is weaker”, I mean this- CO2’s 10.6um is SO effective because almost nothing has any significant transparency to it. So much so that only a handful of exotic materials exist to make lenses for it. A blue laser is not fundamentally different than just blue light. If I had a 450nm blue LED and put a sheet of notebook paper over it, I can still see some of the light through it. If I pointed the LED at a piece of light colored wood, it illuminates the wood because the wood lets some light reflect. If I put a metal housing around the LED and hold it up against the wood so there’s no gap, I can still see a blue glow around that point because the light isn’t absorbed completely, some of it is reflected and spreads around.
10.6um is not like that. You can’t see it, but it’s remarkable in that basically any thickness of anything is essentially totally opaque to it. No only will nothing reflect (except metals), but all the energy absorption is all concentrated at the surface it hits. There’s no gradient where some of the energy is absorbed by the first 0.1mm and some makes it through to the next 0.1mm of depth. That does hinder cutting, the goal is getting the power to turn to heat in the smallest volume to create a high enough temp to turn wood into gas, which ablates that volume exposing more material beneath it. If the beam is more gradually absorbed through the depth, then having a small spot size isn’t exactly equivalent to having that spot size in CO2 because the absorption isn’t all concentrated at that spot like CO2.
Blue ray lasers are also weird in that the material’s absorption is often very nonlinear- paper’s color changes as it is exposed. White paper allows some light through, and it diffuses laterally too. But if the light gets it hot enough, then it becomes brown paper, which is a much stronger absorber, then it becomes black, then ablates away into gas. So how small of a volume that laser energy turns to heat in is not simple to calculate.
Thank you for the additional material. It does explain why I can do some things with a diode laser I should not be able to do.
It is claimed, and appears to be true, that Sculpfun lasers have an exceptional Depth of Focus compares to other lasers. I have seen 40-50mm stated. Personally, I do not see a difference with my 10w machines even when I am 1/4" too high (can’t get that far down). This allows me to evenly laser most curved surfaces like mugs and glass without using the rotary.
Ha, still not too old (78) to learn!
ok had some time to test 5.62mm thick cast acrylic
On a 2" lens, I got 26mm/s. 1mm closer than ideal got 25mm/s, 1mm further than ideal got 21mm/s
4" was much poorer. 17mm/s at ideal focus, 12mm/s 1mm closer,14mm/s 1 mm further
So, the 4" was really inferior on 5.62mm. Easy to explain since it’s 1/4 the energy density at the focal point. But I did expect it to be significantly less affected by being +/- 1mm out of focus. Instead, it seems even more impaired by being off the focal point.
Here’s an interesting test. This is a 68mm deep acrylic block, I just fired for asd long as it took to get through. The upper channel was the 4" lens. The 2" lens initially made a channel about 20mm deep that was narrower and faster than the 4" did. But then it kind of petered out, each mm past than got slower. The time needed to get that last mm was many times slower than the 4". And in that time, the channel that was already there was ballooning out wider.
That’s as bit deceptive because the time the laser was left on with the 2" was much, much longer. The 2" was narrower and faster initially, maybe up to 10mm? Then the 4" pulls ahead, cutting the same width of channel at a most consistent rate. The 4" slowed some as it got deeper but the 2" slowed to a crawl.
Basically the 2" was the hare that was much faster out of the starting gate, but after something like 10mm or 20mm the 4" overtook it as the 2" slowed to a crawl.
The lower 3 were on a 2", fired for a short time, like 1-2 sec.
The upper was done at the focal point.
The next two were about 1 mm and 2mm closer, putting the focal point inside the material. You can see the beam waist form a little funnel where it converges just below the surface.
Also, not labeled, but there are 2 shots above the one labeled 4" which were done on the 4" lens to a partial depth. You can see that the channel is clearly straighter and narrower than those 3 shots on bottom that went to a similar depth.
The cone-shaped things are on the far side of the block, done without a lens.
Great set of results Danny. So the 4" might still be useful for deeper acrylic given enough time. That could be very useful sometimes. And as somebody else mentioned, the 4"could be better for cutting something like shadow foam. I tried some tests recently with the 2" and that was fine up to 25mm, after that the energy had to be increased so much that the lower half of a 30 - 35mm cut was really tapered outwards, so not much good. I’ll try the 4" on the same stuff and see what happens.
Cheers,
David