Fiber coupled laser diode teardown

[blueskull]

I tore down and reverse engineered a FocusLight 8W IR laser diode. This diode is a 915nm, 8W laser diode coupled into a 105um/125um/250um 0.15NA multimode fiber.

https://youtu.be/eD2KMd2FxOo

Still being processed by YT, FHD version should be available in a couple of hours.

[lukier]

Very nice video!

I have a question on the constant current driver. What do you use at these current levels? Do you have something off the shelf or maybe some DIY (schematic?).

I'm asking because I would like to get back to my laser experiments, but this time with IR more powerful laser. I've tried 'etching' black paint painted PCB with 1W 445nm laser mounted on my CNC machine, but it wasn't very good (see here: https://www.eevblog.com/forum/beginners/practical-proscon-of-laser-vs-milling-of-pcb's/msg766269/#msg766269).

Maybe 915 nm 8-25W fiber coupled would be better. I could place the laser elsewhere (with TEC and heatsink) and just ran the fiber to the CNC machine head (I need to think how to couple a focusable lens to the fiber). I might even build a lighter belt driven machine (like cheap 3D printers) to speed things up (CNC with screws is quite heavy and slow).

I have some low power diode drivers from http://odicforce.com/, up to 2A and I also recently aquired ILX Lightwave LDC-3722, very cool device, but only goes to 500mA. I also have a Coherent 100W laser power sensor head, so I could check and adjust the optical power. So if I get a 8-25W laser diode the only thing to solve is a 14A constant current driver with voltage compliance. Maybe I could design something and use a bunch of power rectifier diodes as a "simulated" load of the laser (I always worry I'll cook the laser diode and waste a lot of money, these things are so delicate).

[lukier]

I don't want to ablate copper. LPKF makes such cool machines and you are right, they use pulsed lasers. I want to just burn the black paint that I covered the PCB with, as shown in the link to my old post. Normal chemical etching then.

[lukier]

Or why not use photosensitive paint and do violet/UV mask lithography? Nowadays at <2W power range deep blue lasers are cheaper.

I've tried the photosensitive method many years ago and was disappointed how fragile the process is. From uniformly coating the PCB, to the quality of the mask (well not the case for laser) printout, exposure times, developing solution strength etc. It can produce awesome results, sure (I once even got a speck of dust imprinted in the copper), but requires extreme care and patience and still leads to frustration. That's why I've been doing toner transfer for years. Not as high quality but quite reliable, fast and easy. Also, I would like to minimize the wet chemical processes and mess in general (that's also why I don't like PCB milling).

If you want to do the hard way of charring black paint, then IR laser does exactly what you want.
Still, I think wasting money on fiber laser does not buy much benefits. You need to know that to let light to shine at a perfect angle that can be collimated, you need a high quality fiber cleaver.
If you ever broke the fiber and requires a new splice, you need a fusion splicer (forget about mechanical splicing on high power laser).
I play with fiber laser because I have all of these. If not, think twice before you buy a fiber laser diode. The factory connector is much fragile than you imagined.

Well you might be right, the diode itself is expensive and I would have to spend much more on the splicer and other tools (I don't have any, never did anything with fibers, well maybe except tearing down broken Tektronix OTDR for parts). I was surprised when I was browsing these diodes on ebay that it's just a fiber, no coatings or plastic jacket for some protection/strength.

Therefore it might be more sensible to get a low cost 40W CO2 laser cutter. I think some people managed to burn the paint off PCB with these and I could use it for other stuff as well (acrylic, plywood etc). The problem here is just the size of these machines due to the laser tube length :/ (and ventilation).