The Big Green Thread (I'm gonna regret posting this in the morning)

[mixedgas]

Approximating the KTP/LBO length.....

From Nicolaas Bloembergen's book...

The first order approximation for SHG conversion is

Pout (SHG) = (K^2) * (L^2/A) * P

Where K is a constant
L is the interaction length
A is the cross sectional beam area
P is the circulating power

The A term increases power when the beam is tighter in diameter. Ie a beam waist sets in the crystal.

The K term comes from SNLO and is a constant based on the material.

If you have a short waist, there is no reason for the crystal to be much longer then the waist. If there is no waist, A is about 1 for our purposes, and the crystal needs to be longer. A=1 is a very bad assumption, but its the best I can do for a narrow, parallel, beam.

Now you know why KTP532 and other lasers have a cavity waist at the crystal.

Steve

[LaNeK779]

well, I can't match anyone at the knowledge level and I know it is too soon to start talking about building something

All i want to say is that when the time comes to put "rubber on the road" I'd be happy to chip in to cover some of the costs of that initial first iteration of our PL build

[planters]

Based on the papers you supplied and a few blurbs on Sam's, I suspect a good starting assumption for the OC is 10% transmission. And you might be able to identify the rod lensing by ramp-ing the inter-cavity power and nulling the mirror spacing to produce the same far field.

One challenge will be that when your research nails down a length for the Vandate and the doubler it will be a best guess center of the envelope dimension. The same goes for the OC. Now slope, threshold and beam characteristics as the temperature is adjusted and crystal aligned/tipped can be studied as can 1064 and 532 PO with drive current. But, how to determine if the the fixed components are significantly off? Despite the expense, it might be worth it, depending on the cost, to order three of each of these components in order to bracket an optimal choice.

I'm not trying to escalate this, but this is what I was suggesting when I said that some additional money may have to be spent to optimize the design. It would be a shame to get close and then end up with a so-so result.

a nice little side pumped L cav 2 watter

Interesting that he states 20W IR and 2W output. Under driven? Careless text?

[badger1666]

Me being a bit dumb when it come to this green build stuff and wanting to learn, could someone give me
A brief explanation as to why the crystals need to be a set lenght ? Please
Thanks

[planters]

Me being a bit dumb when it come to this green build stuff and wanting to learn, could someone give me
A brief explanation as to why the crystals need to be a set lenght ? Please
Thanks

They do not have to be a set length as there is no significant self absorption. They need to equal or exceed a minimum length to fully interact with the pump focus and to contain the 1064 waist within the doubler. Not being absolutely certain what these dimensions will be the two crystals (assuming the rod is not a glass or ceramic just to keep the terms neat) then you want to keep them longer with some margin than your best guess. This, of course adds to the cost, but as Steve R. noted, a lot of that cost is in the polishing and coating and so it will not increase proportionally with size.

A consideration will also be that if the optical cavity is not symmetrical such as a semi-hemispherical (concave facing a flat) or the two ROCs of two concave mirrors are not equal then as the Vandate rod increases its + focal power with increased pump power the beam waist will shift longitudinally and must still remain within the doubler. So, both the length of that element and the relative position of the waist within the doubler will need to be considered.

The article by LIU Jun-hi comes pretty close to a cookbook (except no OC%). Easy read.

[mixedgas]

Quote
The article by LIU Jun-hi comes pretty close to a cookbook (except no OC%). Easy read.[/QUOTE]


Are you suggesting that when I write up something for PLublication on Sunday, I use that one as a core citation?

Eric, et al...
I'd suggest downloading PSST at http://www.st-andrews.ac.uk/~psst/ and taking a long, hard, look at the "Lens Duct" function in "Laser Resonators".
That lets us possibly simulate a rod in a cavity if we can measure its lensing when hot.

The trick is thinking in PSST!, as it only lets you enter a linear resonantor with no bends, mirrors etc, except the OC and HR. It also does not let you put in a limiting arpeture, I begged them to put that in, but they were out of funding/time/interest. Its great for ion lasers, but its a stretch for our use. Still, its by laser professionals and supplied by English taxpayers, so what else can you ask for! With the caveat that what looks good in software, usually does not match what we see in the lab. Considering the next possible laser simulator is 5 Grand, we'll use it.

What you cant see in PSST! is the astigmatism in the beam when you fold it. You have to fold it/bend it a least once at these power levels. The fold somewhat helps control the location of the second intracavity beam waist, needed for the KTP.

How to model the astigmatism is covered in this "Classic" from Bell Labs: I don't understand much of it, I'm not a math person, I'm a wrench turner.

" Astigmatically compensated cavities for CW dye lasers "
Quantum Electronics, IEEE Journal of
Date of Publication: Mar 1972
Author(s): Kogelnik, H.
Bell Telephone Laboratories, Inc., Murray Hill, NJ, USA
Ippen, E. ; Dienes, A. ; Shank, C.
Volume: 8 , Issue: 3
Page(s): 373 - 379

I could learn ABCD matricies, if I had the time. Note to others, ABCDs are 4x4 unit array math equations. Used for modeling optics when your poor, dont have a big computer, and fancy software is out of your reach.


This patent is desired reading for open source material.

US 5025446

One other thing to keep in mind is that the Green from the KTP/LBO/BBO does not come out exactly co-linear with the IR beam in one axis. The walk-off angle shows up in SNLO when you run Q-Mix, saving us from having to guess.

Also, if you build one of these, you can expect to see up to half your total green setting back in the cavity some place, with no means to extract it. Physics is a B. some times when considering $ per photon. Under some cases, a 2 watt green is leaking 1 additional watt of green some place where its not easy to get it. If you have a older Melles Griot, you've seen the green two beams coming out, and had to combine them.

Steve

[planters]

Steve R.,

These are nice articles! The laser built by Yihong Chen et al also looks promising. With an 800mm FL OC and 8% transmission they produced a TEM 00 beam dia of 0.56mm and a divergence of 1.74mrad. The M^2 was 1.19! This is about 4 times smaller than the beam product (dia. x divergence) of the typical commercial CN laser. The slightly lower out put power at 11% vs 8% transmission at all input power levels and especially at the higher powers suggests this might be about right even if we run at lower pump powers. This is probably due to lower inter cavity intensities and therefore lower conversion efficiency with the the 11% mirror. Another reason I believe that this is the mechanism is that there is also a reduction of output power with the longer ROC output mirror. This will increase the diameter of the waist within the doubler. Nevertheless, the improvement in beam quality with the longer (800mm) mirror is worth the small reduction in power.

Question: do we look at a 8% mirror and also explore a 6% mirror?

On the modulation front, because we are not cavity modulating (Q switching) there will not be any risk of "over charging" the cavity "capacitor" between output pulses. Furthermore, I would suggest that the thermal effect (lensing) will average out to be similar to the "average" thermal load. In any case I am not too worried that the performance will deteriorate catastrophically. We may need to just build it and test it under modulation.

[planters]

What you cant see in PSST! is the astigmatism in the beam when you fold it. You have to fold it/bend it a least once at these power levels. The fold somewhat instigates the second intracavity beam waist, needed for the KTP.

I don't understand this. What is the purpose of a fold with a radiused mirror? I understand clearly why this would produce a second slightly displaced waist. But, why do they do this? Is this to couple to a birefringent property of the KTP?

The second article I cited just after your post (Yihong), may have oversimplified the layout, but there appears to be no fold at all.

[mixedgas]

Folds make it easier to align. I mispoke a bit there. You CAN do it without folds. However the fold also gives you a nice place to put the wavelength seperator where the beam is narrow and a way to put the waist at a reasonable place on the baseplate.

Brain Fart on my end.

Sorry,

Steve

[mixedgas]

Quote:
On the modulation front, because we are not cavity modulating (Q switching) there will not be any risk of "over charging" the cavity "capacitor" between output pulses. Furthermore, I would suggest that the thermal effect (lensing) will average out to be similar to the "average" thermal load. In any case I am not too worried that the performance will deteriorate catastrophically. We may need to just build it and test it under modulation.[/QUOTE]


It is possible to "Snap" a CW cavity by biasing it just below threshold, storing in energy, and dumping in a big pulse. This is done with microchip lasers to get those rangefinding pulses. However the Qswitched pulse produced is long and weak compared to a real Qswitch in a long cavity such as ours.

Everything I work on over a Watt, pulsed or otherwise, ramps the pump diodes up and down on startup. They also have toggle buttons, allowing you to set a working power and a alignment/tuning power. But I work on Rolls Royce, Ferarri, and Miatta's.

Until we build one, the effects of modulation are just a guess. We can, however, include ovens. I dont know where self heating of the KTP shows up, It does show up at 2-3 watts in the ring lasers in the lab. Its a concern at 4 watts according to the built in software on the 10 watt. . But that is on ultra stable lasers.

Which leads me to the next question, doubler oven or no oven?. QE uses the crystal hanging out on a TE. This means either they dont worry about it much or have a really stable TE electronics.

Also measuring the small electric field generated in the KTP/LBO lets us see when the doubling is peaked. Most of them are piezo materials. Learned about that trick last night. Really shows up on a Qswitched laser. Its a mV or so on a CW laser.

BTW the major dimensions of the laser in Jon's pic are roughly 104 x 144 mm. Just knowing that does not give you anywhere near enough data to clone it.


Steve