If at first you don't succeed....Nd:YAG 3rd Harmonic

[Slow Poke]

Whelp, UV Ablation has been the domain of exacting, well heeled engineers.  Most agree, there is no finer prototyper.  Um....holiday funny!

Cost: Too much for personal use.  Or......is it?

https://www.instructables.com/id/Custom-PCB-Prototyping-using-a-Laser-Cutter/

Desktop laser cutter or CNC mod.  C'mon now --> anyone try?

[Slow Poke]

Not optimal....but:

https://youtu.be/uIIwU29H3E8



Now consider:  Coherent OBIS LB 355nm head  -->  http://lasers.coherent.com/lasers/355-nm-laser

Just about 2"x1" footprint; 5" height --> looks easily mountable

Have not located price; wonder if 50mW is enough - LPKF's 5000 series uses 10-27W laser sources.  That power comes at a  >cough<  premium....but is easily available.

[Slow Poke]

10W -> Getting there:     https://www.laserfocusworld.com/articles/print/volume-37/issue-5/world-news/world-news/diode-pumped-uv-laser-delivers-10-w-at-355-nm.html

Needs moving table....above link was over 17 years ago....






Time marches on and so does R&D.  This year from MKS, across Rt 101 from the In-N-Out Burger south o'the Lawrence Expressway, Silicon Valley, a >45W 150kHz 355nM laser suitable for excellent PCB definition and resolution:   https://www.spectra-physics.com/company/news/mks-launches-spectra-physics-45-w-uv-lasers-with-disruptive-cost-performance  Warning:  Beam stability achieved after 2 hr warmup.

9"x5"x28"....mounting may be an issue....don't see any on Ebay....and call me crazy, but this may replace chemical etch processes for large scale pcb mfg like plasma replaced cleaners.

Can we see a unit for the desktop?

[Spirit532]

Anything pulsed is going to be way above $3000. The cheapest pulsed Chinese 20W Nd:YAG laser is around $3000, look at Raycus and Maxlaser. They output a fiber, that gets piped through a collimator lens block and onto two galvos, then through an f-theta lens and you get a $5000 "metal marker laser engraver 20W CNC yag" or any of those words on Aliexpress and Alibaba.
Pulsed UV lasers are easily over $20k for any reasonable power, but IR can ablate copper.

[LukeW]

So it’s directly vaporising the copper? Any particular reason that frequency-tripled Nd (355 nm) is used as opposed to the Nd:YAG fundamental frequency, or ytterbium fibre laser etc?

If it’s UV laser for photolithography, 405nm diodes are orders of magnitude cheaper and essentially identical to the mercury H-line which is fine for many photoresists when dosed appropriately.

[jeremy]

Yes, if you know of a third harmonic 355nm laser for cheap, I’m all ears! Just the pumping diodes will probably blow your budget Also keep in mind these are Q switched lasers, so the actual instantaneous power is many kW even for a “6W” laser.

[jeremy]

So it’s directly vaporising the copper? Any particular reason that frequency-tripled Nd (355 nm) is used as opposed to the Nd:YAG fundamental frequency, or ytterbium fibre laser etc?

Absorbtion and ablation is better at UV, and I guess the diffraction limit is 3 times smaller? From what I have seen, these lasers are actually often Nd:YVO4 for higher efficiency.

[ajb]

Now consider:  Coherent OBIS LB 355nm head  -->  http://lasers.coherent.com/lasers/355-nm-laser

Just about 2"x1" footprint; 5" height --> looks easily mountable

Have not located price; wonder if 50mW is enough

OBIS heads are simply not economical unless you really need good beam quality, high stability, or one of their unusual wavelengths.  You're looking at somewhere in the $10-30k ballpark depending on power and wavelength, similar for the comparable Oxxius heads.  They are really nice to work with though.  We've looked at using them for a couple of projects at work, unfortunately the customer never moved forward so we didn't get to build more than a prototype.

The rig in that youtube video, with two 405nm diode lasers, is going to be FAR more economical, and probably good enough for a home fab.  (Side note: I'm actually surprised at how well that works, although I can't say I've ever done the math on power requirements for ablating copper.)  An OBIS or similar with near TEM00 beam might get you better resolution (although single-mode diodes can be pretty damn good), but I'm not sure that increased etching resolution is going to be worth a whole lot unless you can also develop correspondingly good capabilities in drilling/plating.

[jeremy]

The rig in that youtube video, with two 405nm diode lasers, is going to be FAR more economical, and probably good enough for a home fab.  (Side note: I'm actually surprised at how well that works, although I can't say I've ever done the math on power requirements for ablating copper.)  An OBIS or similar with near TEM00 beam might get you better resolution (although single-mode diodes can be pretty damn good), but I'm not sure that increased etching resolution is going to be worth a whole lot unless you can also develop correspondingly good capabilities in drilling/plating.

The video is not ablation, it’s exposing the photoresist. ~1W CW is far too low to ablate copper.

[Xnke]

That video is definitely ablation. Do you not see the sparks of copper vapor coming off?

[RoGeorge]

Impressive speed! 

Just to be clear, the laser does not remove the copper.  The white thing is the fotoresist layer after being impressed by the UV laser.  Later, the PCB needs to be etched chemically.  The video here is to show the speed, it does not show the chemical etching.  The chemical etching was shown in a previous video.

[jeremy]

That video is definitely ablation. Do you not see the sparks of copper vapor coming off?

Aside from the fact that the power is simply too low, the only thing I can see which looks like sparks is actually just the beam as an artefact of the rolling shutter of the camera sensor; the beam looks tilted because the laser is moving faster than the camera line scan rate. Copper sparks would probably be green (but I'm just guessing on this one)

Also, if you read the comments, the author of the video does say specifically that this is photoresist only.