Hoping for tight focus

[NoodleSlayer]

Hi gang, I don't know enough about optics to do this on my own, so I'm hoping for some advice. My project has a couple of tight constraints and a bunch of loose ones, so I'm hoping it's feasible.

First off, the goal is to melt plastic in a very small spot, and it's cost sensitive. (Those two factors are the challenges.) I'm hoping for a spot around 1-2 thousandths inch diameter, but 3 or 4 might work. Since I don't know the needed power (that'll have to be discovered experimentally), I figure the cheapest way in the long run is to get a fairly beefy laser and PWM it to determine what lower power unit would work. The idea is to buy one laser once, rather than a succession of more powerful units until I get to the sweet spot. So I've been looking at something like this: 635nm 180mw TTL TEC cooled OEM 12VDC, AixiZ figuring that's probably more wattage than I likely need. So the spot size and cost targets are the bad news.

The good news is the target plastic is black, so any wavelength from IR up through the visible range should work okay. I can handle the electronics fine, so control issues shouldn't factor in. My design will hold a fixed mechanical spacing, so nothing fancy there need be done (no autotracking or anything fancy like that.) The spot does not have to be perfectly circular; elliptical up to maybe 1:2 ratio should work fine. Distance from the front element to the spot can be almost anything, though it would be inconvenient if it were less than maybe 1/4 inch. Aixiz suggested that this lens AixiZ Red Laser Glass Lens 635-670nm, AixiZ would work with their lasers to achieve a small spot, but they didn't have even wild guesses at numbers for feasible spot size, so I lack confidence in that recommendation barring confirmation from someone more technical.

So... can anyone either verify those recommendations, or suggest alternatives, including "oh I could sell you the parts you need"?

Thanks!

 

[Cyparagon]

Please tell us more about your project... specifically, why this melting is needed.

 

[slymartins]

How many microns are you looking for, for the spot width? At those ranges, it may be best to refer to measurements in microns(micrometers) as I believe it's more of an industry standard than thousands of an inch, and it may help with your online research too.

I've been reading a bit into this myself recently, and from what I was able to find, the best solutions may rely between using razor blades for physical shaping of the beam, coupled with a good GRIN lens (which focuses much, much better than a regular aspherical). Would be interested to hear from other folks on it, though. Just my 2 cents in case it helps.

 

[NoodleSlayer]

Yes, okay, 0.001 inch = 25.4 microns, so while I'd really like 30-40 microns, 80-100 might work. And I understand the edge will be a bit fuzzy, so any measurement of spot diameter is a bit imprecise by its nature.

The goal is to fuse laser toner particles, as a slightly new variation on making of a resist pattern for PCBs. Toner particles being most commonly around 10 microns in diameter, and the unfused particles would be several layers deep, "a few particles wide" seems a reasonable goal for the width of the meltie-path. (FWIW, this has been done before, but with a stronger, CO2 laser and with a fatter beam, resulting in of course rougher resist patterns.)

Thanks!

 

[lazeristasUVISIR]

I'd really like 30-40 microns, 80-100 might work.

Depends on the laser diode. Look for a single-mode LD. As your material is black the required power might be low.

Get a lens with a longer focal length like 10-15 mm and NA~0.3, and focus at ~100 mm, you should enjoy 50-um spots then.

 

[pschlosser]

You want to use visible light, focused to 100 microns, is that right? To do this, you’ll need optics to first expand the beam, and then focus it down to a tiny spot. But that, of course, makes the (depth of field small and) focal length critical. Meaning you’ll have use an XYZ table to move either the target or the beam and have some rather tight tolerances on positioning to maintain the focal length.

Although toner is a polymer (a plastic) it is, by some definition, a wax. When heated, it melts and bleeds at the microscopic level. It seems your proposed approach is to selectively cure tiny areas of a sheet of toner, hoping the toner adheres to the substrate, and the uncured toner is removed, is that right? I think you will find the bleeding difficult to control. Heat is not your friend for tight tolerances, because of its longer wavelength and its propagation into the medium targeted.

A laser printer, by contrast, makes no attempt to use light to heat or melt toner. In a laser printer, and many digital CMYK printing presses, a laser is used to alter the static charge of a drum. The ink or toner either sticks or doesn’t stick to these differently charged areas of the drum and an image forms. The image on the drum is then transferred to the paper (substrate) where heat is used to fuse (melt) the ink or toner to the substrate.

Ultraviolet is the preferred wavelength to blast material away. Many pulsed varieties exist in the industry to make very clean and tiny holes and etchings into all kinds of materials. Google “fempto laser” and you’ll see what I mean. I visited a company near me that demonstrates this by etching their company logo into the head of a match stick without, of course, lighting the match.

Printed circuit work (at the small scales you describe) that use laser light are using the UV wavelengths and pulsed technology to blast away unwanted material.

If you are attempting to use toner to make a mask, I recommend AutoCAD and a laser printer to create your mask onto a medium such as paper and then transfer the mask onto the substrate targeted. You can easily get 600 DPI. Higher, if you locate a better printer.

 

[slymartins]

Great info, pschlosser. Thanks for sharing.

 

[NoodleSlayer]

Depends on the laser diode. Look for a single-mode LD.

Excellent, this sort of practical specific is very helpful.

As your material is black the required power might be low.

I'm hoping so. My thought is prototype 1 gets a heavier laser than needed, I'll reduce average power by pulse width modulation to discover what's truly needed, then prototype 2 would (assuming all went well) use a smaller, cheaper, lower power module, sufficient to the task but more economical.

Get a lens with a longer focal length like 10-15 mm and NA~0.3, and focus at ~100 mm, you should enjoy 50-um spots then.

Okay, I'm off to discover what numerical aperture is all about. Again, such specifics are very helpful. I'll reply to the next message when I know a bit more.

I'm guessing a longer focal length will give me a longer waist, easing the task of maintaining correct distance?

Thanks for the suggestions, I'll be back in a little bit, most likely w more questions! :yh:

 

[NoodleSlayer]

You want to use visible light, focused to 100 microns, is that right?

Close. I'm agnostic about wavelength, don't even care if it's visible. While I plan on manual, static focus, the laser need not be visible to do that. I'm planning to use one of the $30 chinese usb microscopes (or a modified webcam, which is pretty much the same thing) to do my looking for me. I'll watch what I'm doing on a pc screen, so even safety glasses, while desirable, should actually be redundant. CCD devices have some response in the IR typically, and if I were to go UV, then a fluorescing target (dayglo type paint) isn't hard to arrange.

100 microns is at the top end of my desires; I'd prefer 1/2 or 1/3 that, but we live in the real world, so we'll see what can be managed.

To do this, you’ll need optics to first expand the beam, and then focus it down to a tiny spot. But that, of course, makes the (depth of field small and) focal length critical. Meaning you’ll have use an XYZ table to move either the target or the beam and have some rather tight tolerances on positioning to maintain the focal length.

I intend for the Z axis to be manually adjustable, and to hold static settings. How well I manage that, of course, depends on how good a mechanical design I come up with. That's my responsibility, of course.

That does open up an area in which I'm rather confused. Laser diodes have large divergence, so need "collimating" (beam paralleling) convergent lenses, widely and cheaply available. But if the diodes have different divergences, how are a tiny variety of lenses sufficient, even for the low performance goal of a handheld pointer? I'd think many focal lengths would be needed, yet that parameter is rarely even specified by sellers?

Another point of confusion: if the commodity lenses achieve something pretty close to a parallel beam of 2-4 mm diam, why would that need to be diverged before re-converged to a focal spot? Why couldn't that be simply fed to a convergent lens directly? I seem to be missing something here, possibly so obvious to others they don't routinely mention it.

Although toner is a polymer (a plastic) it is, by some definition, a wax. When heated, it melts and bleeds at the microscopic level. It seems your proposed approach is to selectively cure tiny areas of a sheet of toner, hoping the toner adheres to the substrate, and the uncured toner is removed, is that right? I think you will find the bleeding difficult to control. Heat is not your friend for tight tolerances, because of its longer wavelength and its propagation into the medium targeted.

Quite right; some experimentation is needed in this area. A fellow I know has done this, but he used a 40W CO2 laser, dialed down to he-doesn't-know-what power level. It's also coarser than I hope to accomplish, though again, he doesn't even know exactly what his gear can accomplish. It's my intention to do a more delicate and precise job of it, using a lower powered laser, wildly different mechanicals, and (sadly, as it's outside my expertise) more carefully selected optics.

A laser printer, by contrast, makes no attempt to use light to heat or melt toner. In a laser printer, and many digital CMYK printing presses, a laser is used to alter the static charge of a drum. The ink or toner either sticks or doesn’t stick to these differently charged areas of the drum and an image forms. The image on the drum is then transferred to the paper (substrate) where heat is used to fuse (melt) the ink or toner to the substrate.

This is a cogent and correct description of the operation of a laser printer. I happen to understand this, but mentioning it in case I were under some misapprehensions was a good idea.

Ultraviolet is the preferred wavelength to blast material away. Many pulsed varieties exist in the industry to make very clean and tiny holes and etchings into all kinds of materials. Google “fempto laser” and you’ll see what I mean. I visited a company near me that demonstrates this by etching their company logo into the head of a match stick without, of course, lighting the match.

Printed circuit work (at the small scales you describe) that use laser light are using the UV wavelengths and pulsed technology to blast away unwanted material.

UV the accepted freq for ablative processes is news to me, but good to know. Should I go ablative, that would be an entirely different project, one I'd much prefer not to pursue, as it'd be much harder to keep the lens clean, nasty stuff would be emitted by the device requiring much better ventilation, some sort of feedback would likely be needed to ensure we don't remove the mask and then start blasting away the copper, and I expect the whole process would run an order of magnitude or more slower. It does offer the prospect of extremely fine work, though.

If you are attempting to use toner to make a mask, I recommend AutoCAD and a laser printer to create your mask onto a medium such as paper and then transfer the mask onto the substrate targeted. You can easily get 600 DPI. Higher, if you locate a better printer.

People are doing that, sometimes with very good results. Printing to paper, then heat-transferring the toner to a PCB is sensitive to printer make and model, toner type, paper type, printer settings, etc all of which people do share on forums devoted to the approach. They also have found that clothes irons work sometimes, but they usually modify heat laminators, run board and paper through several times, adopt various procedures to achieve and keep alignment of the pieces, etc. Pinholes are a problem, mainly in large masked areas, and several fixes for that have been tried, some with success. One of the better approaches I've heard of uses a vinyl sheet, as it seems to release the toner more easily than paper. While that approach does work, once fine tuned and done with great care, I hope to make this approach simpler to run and more reliable and precise.

Your comments about melting and flowing triggered a thought: My expectation is to achieve some fusing, though perhaps not enough for final etching. (That's not a given, as etching fluid has a surface tension, so a mask might not need to be perfect to be good enough.) But suppose that after laser fusing and brushing away the loose powder (drug store fluffy cosmetics brush?), what remains may be porous and insufficiently adhered to the copper. Okay. It may be that simply running the board under a heat lamp as a second step might reflow the toner adequately. That would be a second step, but simple and cheap enough to not be a problem. I hadn't thought of that before.

For the time being, my focus (no pun) is on selecting a module and lenses to achieve power and precision I hope sufficient to try this approach. I don't yet know enough to select from the affordable options a likely combination. Advice in that area would be very welcome, whether from a vendor or not.

Thanks for the reply and the time and attention that went into composing it!

 

[NoodleSlayer]

What sort of divergence angles can we typically expect from a single mode diode, without optics, eg coming straight out of the diode can? Should I assume I'm most likely to find a suitable diode in the 600 nm range?