Laser Combination: Questions & Answers for PCB Etching

[adaminc]

Alrighty, I have 2 simple questions about Laser Beam Combination.

I was wondering if I could take multiple Laser Diodes and direct them all at the same point and have the combined energy of the lasers at that single point. Would phase cancellation be an issue?

If that doesn't work, I was thinking about using 50% silvered mirrors inorder to combine the beams, and I was wondering how much power loss is there from a 50% silvered mirror, is it 50%? or is that just the amount of light it let's through.

The reason for all this is I am doing an experiment in laser pcb etching, however not actually etching the copper, but burning/charring a thin layer (like paint) off the top, which I would then wash off, or blow off with air, then throw into an etching solution.

Also, could any of you recommend a good book on lasers?

 

[mgb_phys]

You can focus multiple laser beams onto a spot. Just put a set of them side by side and have a row of small mirros at45deg so that the beams just pass the previous mirror and come to a point.
Another common technique in high power lasers is to take a bunch of fibres from the lasers and put them together into a bundel to form a single round source of light you can focus. This is especially common with diode lasers which are normally flat slab strructures that are otherwise difficult to focus.

The main problem with your PCB etching is probably wavelength.
High power tends to come with longer wavelengths (if you want to burn through thick steel 10um CO2 lasers are common)
PCB photoetch tends to be UV sensitive - uv lasers are a pain.
Can you use a visible/near IR sensitive imulsion and use a regular red laser or a 905 near IR diode laser.

 

[adaminc]

When you mean use a row of small mirrors, do you mean so that the laser will bounce off one mirror and pass through the next mirror to combine with the beam on that mirror? or do you mean just have a bunch of mirrors that are redirecting the beams into a specific spot?

The laser diodes I intend on using are 250mW DVD Burning Lasers, they are visible (Red), however since they are so cheap, I intend on doing multiple tests to see if I can PWM them to ~500mW, and I was wondering if by combining say 2 250mW lasers and focusing them onto the same point if it would be roughly 500mW worth of power.

As for the actual layer I will be burning, I will probably try starting with regular black matte paint painted ontop of the copper layer.

 

[bsimmo]

It not so much the power of the laser that is important as the wavelength you use. No point using a CO2 laser if it cannot be absorbed effectively.

Esencially you're trying to remove a resist of the top of the Copper (this practice is already used for a long time.

You would want a wavelength that copper is not removed at but your resist is. A plastic coating such as PMMA would be nice (you can spin these on). But it depends on knowledge of laser machining ?

btw high power laser diodes are just a collection of the low power laser's stuck together.

250mW @ ~650nm should cut thin tape without a problem
Power loss from a mirror is ~equiv to what get through, so check what get through for 650nm.

 

[GT1]

You can focus multiple laser beams onto a spot. Just
The main problem with your PCB etching is probably wavelength.
High power tends to come with longer wavelengths .

Why high power comes with longer wavelengths ? I thought that the shorter
wavelengths are more energetic.

 

[mgb_phys]

Why high power comes with longer wavelengths ? I thought that the shorter wavelengths are more energetic.

They are which means you need a material with a higher energy level to generate, which means it is harder to create a large population inversion.
It's mostly a matter of engineering - it's easy to create high power infrared gas lasers so this is the area that has had a lot of industrial research and where a lot of products are available.

The main difficulty with the DVD lasers is that the diode will have a slab structure and contain a built-in micro lens to reshape and focus the beam. It might be tricky to colimate the beam in order to run it through a whole table of optics.

 

[Math Jeans]

http://www.repairfaq.org/sam/lasercc2.htm#cc2cc20"

 

[seycyrus]

Alrighty, I have 2 simple questions about Laser Beam Combination.
I was wondering if I could take multiple Laser Diodes and direct them all at the same point and have the combined energy of the lasers at that single point. Would phase cancellation be an issue?
QUOTE]

Yes, I believe it is an issue. The beams might be selfcoherent, but are not coherent with each other.

There is some interesting work however, in multiple beams in a single fiber optic forming a self organized critcal coherent state.

 

[Claude Bile]

I was wondering if I could take multiple Laser Diodes and direct them all at the same point and have the combined energy of the lasers at that single point. Would phase cancellation be an issue?

Kind of, the beams will interfere, giving a fringe pattern. In this case having highly incoherent laser beams would be advantageous.

The reason for all this is I am doing an experiment in laser pcb etching, however not actually etching the copper, but burning/charring a thin layer (like paint) off the top, which I would then wash off, or blow off with air, then throw into an etching solution.

Do you need to pulse your lasers for this type of application? If so, then that introduces a heap of challenges when it comes to synchronising your lasers and getting to fire them in a coordinated fashion so each pulse arrives at the target at the same time.

Claude.

 

[adaminc]

I don't have to pulse them, it would be nicer since I could over power them, but its all going to be controlled from a microcontroller, so I am hoping that sync issues can be easily solved just by tweaking values. But if it does become too difficult, then I will just run them constantly.

 

[adaminc]

I was thinking about what everyone has told me and I am wondering if you can easily combine beams of different wavelengths without interference, e.g. a 650nm and a 720nm or 808nm beam?

If so, I was also thinking that if I could combine 2 or 3 of those different wavelengths, and have two sets of these combinations, and alternate between the two, then I would essentially be creating more heat energy at the spot, more easily.

So instead of 2 or 3 beams of 1 wavelength hitting a spot with a 50% pn/off duty cycle, I could have 2 sets of 2 or 3 beams hitting a spot, each set hitting when the other is off, and this would create more heat energy, I would think.

 

[bsimmo]

Can I just say.

The first thing I would do it get your DVD laser and just power it using (search the web and you'll find example of making them into pens with the simple collimating lens)

Interference may be a problem but then you'll more than likely find it isn't (best guess but then just tweak your laser distances to compensate if you do find it acting like an interferometer ;-) oh and the main problem you have with different wavelength is your focus will be in different places and will behave differently to lenses.)

Why not try the 532nm DPSS lasers...

Until you try the single diode, you are thinking to much into it without testing the basics. It's not the *energy* that matters is the amount of energy at that wavelength that can be coupled into the material you need to process, (the black pen but not touch the copper).

But with PWM how do you think this will increase a laser diodes power?
Laser diodes have a constant wave output, if you modulate them say 50% then you half the average power, effectively halving the energy output. If you increase the driving voltage to increase power, then just use decent cooling (TEC's for example) to stop the wavelength drifting and the shape broadening.

Go to far and they are not going to last very long either ;)

 

[Claude Bile]

I was thinking about what everyone has told me and I am wondering if you can easily combine beams of different wavelengths without interference, e.g. a 650nm and a 720nm or 808nm beam?

If so, I was also thinking that if I could combine 2 or 3 of those different wavelengths, and have two sets of these combinations, and alternate between the two, then I would essentially be creating more heat energy at the spot, more easily.

So instead of 2 or 3 beams of 1 wavelength hitting a spot with a 50% pn/off duty cycle, I could have 2 sets of 2 or 3 beams hitting a spot, each set hitting when the other is off, and this would create more heat energy, I would think.

You could use multiple wavelengths, but it introduces complications. Different wavelengths will affect the substrate differently (for a given power).

My advice would be to cut down your parameter space - use a single laser, only use multiple lasers as a last resort.

Claude.

 

[adaminc]

I am going to use just a single laser, and I have a line on getting a 2w 808nm Laser Diode for $70 each (I don't expect them to be top-notch or anything), that makes this NIR light, which might or might not be visible, anyways, I was wondering if I will need a collimating lens or a focusing lens if the diode will be within 1.5" of the PCB (there will be a thin piece of glass between the laser and the work area to protect the diode). I guess it depends on the spec's of the diode, doesn't it.

Spectral width(nm):≤5
Beam divergence(deg):12/40
Emitting area(μm):200x1

I am also wondering what the 12/40 means, I know it means 12 degrees and 40 degrees, and it has to do with the beam divergence over a distance, is it something like 12 degrees at the aperture, and 40 degrees after like 1mm or something? Any info that could help me fine-tine the resolution of etching would be great.

 

[bsimmo]

808nm should not be visible, other than as a very dim glow at the target.
If it is just a diode without a collimating lens then yes you will (depending on your area/neatness requirements, look at the divergence of 12/40 degrees* (vert/horiz probably). You may be able to focus with this as well... Don't forget to get some IR filtering safety glasses, IR lasers to the eye are not good, even reflected.

* it means the beam spread, normaly you see it in mrads
40degrees is wider than some spotlight (the GU10/GU5.3 downlighters)


Melles Griot always had some nice info

http://www.mellesgriot.com/products/technicalliterature.asp