KTP and DPSS

[Hiemal]

Hey all.

I know that inside of green laser pointers there is a DPSS (diode pumped solid state) system.

There's an 808 nm ir diode that goes through a crystal to make 1064 nm, and then through a KTP crystal to halve the wavelength.

From what I've read, KTP works all the way down to 994 nm...(upper frequency range I didn't find out)

Is there any way to possibly make a 500 nm DPSS system using this idea? (or any wavelength in that region?)

If there WAS an IR diode in between 980 nm and 1064 nm that is... How does the crystal that bumps the 808 nm up to 1064 work? (is there any other frequency that could possibly be used?)




And to those who say alignment and mirrors and cavities,

KTP crystal doubling a pulsed Nd:YAG laser - YouTube seems to work pretty well considering he's just holding it there with a pair of pliers.


Just some thoughts of mine, and well, I apologize if any of them are kind of stupid.

Thanks for any answers! :wave:

 

[Sigurthr]

And to those who say alignment and mirrors and cavities,

KTP crystal doubling a pulsed Nd:YAG laser - YouTube seems to work pretty well considering he's just holding it there with a pair of pliers.


Just some thoughts of mine, and well, I apologize if any of them are kind of stupid.

Thanks for any answers! :wave:

Mirrors are actually formed by the coatings on the crystals. A cavity is a gain medium between two mirrors. If you coat two opposing sides of the crystal you now have a cavity.

You would need a pump source (for the KTP) at half the frequency of your target wavelength as well as specialized coatings on the KTP to form the resonant cavity at the intended wavelength. Coatings need to be HT to the pump and HR to the target wavelength on the entrance side and HR to the pump and OC to the target on the exit side.

In the video is a coated KTP crystal designed for 532nm output. If it were not coated or not designed for 532nm you would not get any lasing, but maybe some regular light emission.

 

[rhd]

There are people on here with a better technical understanding of the precise scientific complications, but I'll explain my nutshell understanding of why this isn't so fantastically simple.

- The actual frequency doubling process, the point at which 1064 becomes 532, requires a VERY narrow bandwidth to work. If your KTP was meant to double 1064, and you gave it 1065, it wouldn't work.
- When 1064 comes out of a crystal being pumped by 808, that 1064 is basically rock solid at 1064, even if the pump wasn't quite 808 (maybe it was 811, or 807, etc)
- This is important, because it means that you can use something like a diode laser, that fluctuates in wavelength a LOT (with heat, and tolerances from diode to diode), yet still get the very narrow bandwidth 1064nm needed for frequency doubling.

This is why "directly doubled" 1064nm diode to 532nm setups are fairly difficult to achieve. They're out there (like the Corning G-1000s), but it's MUCH more difficult to keep a 1064nm diode locked on 1064nm (tightly enough) than it is to get that 1064nm from a crystal, pumped by an 808nm diode that can wander wherever the hell it wants within a few nm +/- of 808.

I've spent a lot of time trying to figure out if a portable 488nm handheld, of any sort, was achievable (at a cost that was practical). In the process of this inquiry, I read this:
Blue-green light generation using high brilliance edge emitting diode lasers - Jechow - 2010 - Laser & Photonics Reviews - Wiley Online Library

It's worth giving it a read. If nothing else, you'll gain an understanding for why it is so difficult to directly double diode sources.

 

[ElektroFreak]

- The actual frequency doubling process, the point at which 1064 becomes 532, requires a VERY narrow bandwidth to work. If your KTP was meant to double 1064, and you gave it 1065, it wouldn't work.

You're pretty much on about everything but I don't think this is completely accurate. According to Sam's FAQ, KTP is generally good to double down to 500nm, not 100% sure about the upper limit but it can be used to produce red light.. You put any wavelength within those limits in and you get doubled output. Most KTP isn't mirror coated, only that used in DPMs IIRC. Any AR coating will be comparatively broadband.

 

[Trevor]

You're pretty much on about everything but I don't think this is completely accurate. According to Sam's FAQ, KTP is generally good to double down to 500nm, not 100% sure about the upper limit but it can be used to produce red light.. You put any wavelength within those limits in and you get doubled output. Most KTP isn't mirror coated, only that used in DPMs IIRC. Any AR coating will be comparatively broadband.

KTP is, however, absolutely brutally efficient at 1064nm. Less so at other wavelengths, if my memory serves.

Trevor

 

[Hiemal]

Would it be expensive to get an AR coated KTP crystal? The ones I've seen on ebay are rather pricey, and it was making me wonder if it would be possible to get one somewhere else cheaper...

It could be fun to experiment with!

 

[ElektroFreak]

KTP is, however, absolutely brutally efficient at 1064nm. Less so at other wavelengths, if my memory serves.

Trevor

Efficiency is highly variable, you're right.

 

[rhd]

You put any wavelength within those limits in and you get doubled output.

Wait, what? There's NO WAY that could be true.

If I just shot a 1208nm diode into KTP, there's no way I'm getting a 604nm beam out.

 

[ElektroFreak]

Why not? 671 DPSS? KTP. I'd imagine if you're direct doubling at a point where the efficiency is lower you'd need to be more conscious of power density and beam specs. Not sure if you could direct double that way or not, but if you found a laser crystal that gives you a good line at 1208nm you'd get something out at 604nm.

 

[Cyparagon]

the point at which 1064 becomes 532, requires a VERY narrow bandwidth to work. If your KTP was meant to double 1064, and you gave it 1065, it wouldn't work.

You're pretty much on about everything but I don't think this is completely accurate.

Relevant, my defective pointer:

 

[ElektroFreak]

How cool is that! Pretty neat defect..

 

[Hiemal]

Then there IS a way to get other wavelengths...

I'm going to try and get KTP.

This is starting to sound pretty neat.

That also makes me wonder, do yellow laser pointers use KTP? I think I may have found a way to make a homemade yellow pointer...

 

[ElektroFreak]

They might, but 594nm yellow uses a different process, sum-frequency mixing which can still be done with KTP. Basically a dual-line emission is stimulated in Nd:YVO4 or Nd:YAG (can't remember which off the top of my head) by using mirror coatings that are reflective at two wavelengths and both wavelengths are passed through the non-linear crystal and summed, resulting in 594nm. I had all the data on hand at one point but I can't seem to find it now. If I find the links I'll post them.

 

[Hiemal]

Do laser diodes between 994 nm and 1300 nm exist?

I'm trying to find some on ebay..

 

[ElektroFreak]

Oh yes. Lots.. But directly doubling a laser diode is very hard, at least doubling that results in any kind of usable beam.

 

[Hiemal]

What if the KTP was AR coated?