So, you're sitting there reading some auction on eBay for a piece of YAG saying it's easy to pump it with 40 watts of IR and get 10 watts of green from a simple cavity. Well, unlucky for you, it's not quite that simple. There are a lot of things you have to understand before you go off buying a expensive diode and crystals hoping to get several watts.

First thing you need to understand is what is actually happening in the cavity that is making the green light. There is a lot of complex stuff working together to make a laser work. There are also several different types of cavities you can use to achieve your goal, some are easier to build then others, but have drawbacks.


Cavities
The first thing I will talk about is the cavity layout. There are 4 main types I will talk about, while there are others, they are not as suitable for a high power green laser. These types are: Linear, L-Fold, Z-Fold, and Ring cavities.

Linear Cavity:

Linear cavities are the simplest of all laser cavities, they consist of 2 or 3 mirrors depending on what you want to do. For a simple IR laser, you will need a set of 2 mirrors, a high reflector, or HR for the rear mirror, and a output coupler, OC for the front mirror. The goal of the cavity is to allow for more gain. When you have no mirrors, you get a maximum of 1 complete pass through the gain medium, this gets you very low power. When you put on one mirror, the light bounces back giving you a maximum of 2 passes through the medium. When you add and correctly align then second mirror, you greatly increase the gain of the medium. For any reasonable laser system, you will require 2 cavity mirrors.

When choosing your mirrors, you also have to deal with the curvature of the mirrors. There is an ideal layout of the mirror's radius of curvature vs. cavity length. For simplicity sake, let's assume we want to build a semiconfocal resonator, the easiest to align and the most thermally stable. Your HR mirror would have an infinite focal length, AKA, flat, while the output coupler would have a radius of curvature that is two times the length of the cavity. This means that the beam waist will be located on the output coupler.

The next thing you have to consider when choosing your mirrors is the way you want to pump the system. Assuming you are using an 808nm diode system, you can either side pump the gain medium or end pump the gain medium with either a focused fiber coupled diode or focused direct diode. The easiest way of pumping in my opinion is to use a fiber coupled diode with a focusing adapter on the end. If you choose to end pump the medium, you will need a high reflector that is both high transmission at the pump wavelength and high reflecting at the laser wavelength. If you choose to use side pumping, then you will need an exposed side of the gain medium that matches up with the length of the emitter on the diode you choose. Side pumping is generally only usable for a large emitter array and not a single emitter C-Mount or canned diode.

The final thing you need to consider when choosing the mirrors for your cavity is the reflectance of the output coupler. With perfect mirrors and no looses internally in the cavity, the intracavity power would be infinite, but due to the fact that mirrors with perfect reflectance do not exist and there are always losses in the cavity, you will always have losses. The transmission of the output coupler is considered a loss of the cavity. For a CW laser system, you want a much higher reflectance output coupler because the laser is always running and thus is always amplifying the emission of the gain medium. Typically, for a DPSS laser, the output coupler is in the range of 95-98% reflectance. For a high power system you want a higher reflectance. For most applications 95% will work perfectly, however will not be the optimal choice.

For a non doubled laser, you will want your cavity to consist of 2 mirrors and the layout will be as follows:
OC@1064 Gain Medium HR@1064

For a doubled cavity with an intracavity frequency doubling crystal:
OC@532/HR@1064 Doubler HR@532/HT@1064 Gain Medium HR@1064

L-Fold:

An L-Fold cavity is the easiest folded cavity to understand. The goal of it is to prevent any 532 from getting into the doubler crystal. They usually do not have as much efficiency as a linear or ring cavity, but they can handle much higher powers and have an artificially longer cavity length due to the folded design.

With a L-Fold, you will have 3 mirrors: 2 end mirrors, and a fold mirror. Both of the end mirrors will be flat while the fold mirror will be curved with a ROC 2 times the length of one leg of the cavity. Essentially, you will be making 2 semiconfocal resonators with a common mirror.

A L-Fold is only used for a doubled laser system, one leg will contain a gain medium while the other leg will contain the doubler. The light will come out from the fold mirror. You want all of the 1064nm light to stay within the cavity so you will have all 3 mirrors be high reflectors of 1064nm light. The end mirrors on the laser should be both a high reflector at 1064nm as well as 532nm. It should be flat. The fold mirror should be a HR at 1064nm and OC @ 532nm.

L-Fold cavities can either be end pumped or side pumped as before. With these types of cavities, it is common to use side pumping.

Z-Fold:

A Z-Fold cavity is very similar to a L-Fold cavity, it is done for the same reason, to artificially extend the length of the cavity. In a Z-Fold, the 2 fold mirrors are curved with 2 different ROC. The shorter leg of the cavity usually has the doubler, while the longer leg has the gain medium. The interconnecting part between the fold mirrors has nothing in it. If the ROC on the short leg is 1, then the ROC on the long leg would be 2. The end mirrors would still be flat.

All of the mirrors would be HR@1064 and 3 would be HR@532 as well, these would be the end mirrors, and the short ROC fold mirror, the long ROC fold mirror, would be HR1064/OC532.

Ring Cavities:

I am not going to talk too much about ring lasers as they are more advanced than the normal hobbiest can assemble, they are used when long coherence length, high gain, and stability is required.


Now onto the gain medium. I am assuming everyone will be using either Nd:YAG or Nd:YVO4 as the gain medium.

YAG vs. YVO4

So most people know about Neodymium doped yttrium aluminum garnet, or Nd:YAG, it is a simple material that is easy to obtain and emits on several wavelengths, primarily 1064nm. It has a relatively high damage threshold and is great if you want a pulsed laser.

Nd:YVO4 is a better choice for high power CW green or 1064nm lasers, it has a higher threshold level, however it has a higher gain and a higher damage threshold meaning it makes high power levels possible.

Doping

Doping is another very important factor in the laser design. With a pulsed laser, you want a high doping level so as to get the most gain in the short number of passes through the gain medium. Too high, and you will get reabsorbtion which can destroy the crystal, too low and you do not get the maximum gain. In a CW DPSS laser, you do not need to deal with the problem of pulsed laser only getting a few passes through the medium and thus you can use a much lower doping to prevent reabsorbtion, yet still provide adequate gain for the number of passes.

For most high power DPSS systems, 0.3-0.5% doping is ideal, much higher and you will not be able to get much power without damaging the crystal.

Pump Lasers

Pump lasers are typically where DIY DPSS projects start, you get a great deal on a 40 watt IR diode which the seller claims is 808nm and perfect for DPSS. While this may be the case, in many instances, they are not suitable. For a laser diode to work in a YAG or YVO4 based DPSS system, you want the wavelength to be centered at 808nm, not 809nm or 807nm. While you can cool the diode to get it to a specific wavelength, this adds another level of difficulty and is difficult to do without a spectrum analyzer or spectrometer.

Diodes come in several packages, the fiber coupled arrays are usually very easy to use and can be used for end pumping without much issue. They can be located away from the resonator. Raw mounted diode bars can be used to side pump materials easily.

Stable Construction

This is another place where people suffer, you must use a very stable adjustable construction method. Glue and wood does not work, you will need a thick metal plate as a base and at minimum 2 adjuster kinematic mounts. The best things to use are 3 adjusters as they offer more versatility. All cavity optics must be bolted down to the base plate along with crystals and doubler.

The Doubler

This is another place where people struggle, in order to get green, you need a doubler crystal, it can be either KTP or LBO, however, for most applications, KTP is preferable due to the smaller size requirement and temperature requirements. In order to get efficient doubling, the crystals must be heated. For most KTP, the optimal temperature is 50-90 degrees Celsius. They must be kept at a constant temperature. This is usually done by means of a heating oven. This oven must have some sort of active temperature control and monitoring system as well as a backup batter to prevent the temperature from dropping too much should the power to the oven be lost.

In Conclusion

I think that covers most everything people get stuck at when planning a DIY DPSS system, it is a complex and expensive project and not something to jump into until you fully understand the cost and requirements.

If there are any obvious errors or things you would like me to add, please let me know.

This is a great writeup! Extremely thorough, definitely should be stickied.

I want to add that this complexity shouldn't deter anyone from trying DIY DPSS construction. Simple, linear DPSS lasers are great fun on their own, and are easily within range of the hobbyist in terms of construction. The more complex cavities are outside of the average hobbyists skill set and equipment to construct, but even they are not impossible if you have built up the knowledge and equipment over time. Like almost anything worth learning how to do, you've got to crawl before you walk. Just don't shoot too high if it's your first time attempting DPSS. You'll most likely end up wasting a good deal of time and money only to give up out of frustration.

Thanks for the sticky!!!

Very nice !

This'll come in handy for me actually

Thank you !

This thread was needed a LONG time ago. Now I can be lazy and link to this thread instead of trying to explain why somebody's 10W green laser project under a pathetic budget won't work. Nice writeup.

I feel smarter after reading that, thanks. Nice writeup

Thanks guys! If you can think of other informative threads that need to be done, let me know.

Great job explaining a difficult subject! Now on to the ring dye laser writeup. Just kidding that is very complex too.

Exactly what my hungry brain needed.....

This is the kind of information I was hoping to find on this forum.

Something on tips and tricks would be good.

For instance, for the DPSS bits, alignment is critical. A trick I read about in a paper on gas lasers, is to sandwich an optic between two metal discs with holes in them, using an O-ring of silicone rubber between the optic and metal. A set of three screws with nuts will let you compress the silicone rubber and thereby shift the angle of the optic slightly. It will need a bit of realigning on occasion, but it's a lot cheaper than a precision mount.

Microscope objective glass and cover glass is good for alignment, and the objective glasses are thick enough to give a little bit of heat sinking. If working with Nd:YAG, they can also work as a Brewster plate if you fiddle a bit with the alignment. Cooling a crystal can be a bit tricky when pumping it from the end face, but an objective glass is better than nothing (at the cost of 4% of the pump power). A sapphire window from EO will give more heat sinking, and the small ones are fairly cheap. You lose a bit more pump power with sapphire, about 20%, but the heat sinking is almost as good as metal.

KTP crystals are cut at roughly the right angle, but to improve the alignment to get better efficiency, you can use Torr Epoxy to mount them in an Aixiz lens nut and then use that to shift the alignment slightly. A big wheel attached to the nut lets you make finer adjustment than doing it by hand. Alignment is, again, one of the keys to efficiency. This approach will not work if the beam diameter is larger than the crystal size (which it should never be in any case, due to diffraction and the like).

If you are using Nd:YVO4, it must be aligned properly to the diode in order to get the best efficiency. This can also be done by abusing an Aixiz module, provided you cut off the back side of the head so that you're left with the threads and the lens nut. Just drill out the nut and glue the crystal to it.

Anyway, those are some thoughts.

The 20W and 40W pump diodes require more than just tips and tricks, so an article on how to use them properly might be good. I know there are many options, ranging from the simple (i.e. get a square crystal and just mount it in front of the emitters at point blank) to the ridiculously complex.

Some tips on good parts sources might be an idea.

If someone on the boards has the equipment, I think it might be nice to sell a few bits and pieces here. For instance, the O-like output couplers require a larger lens nut. I also haven't seen many places you can get a crystal holder if you're not specifically buying their crystals (and often in large quantities), so a simple holder plus O-ring and retaining ring might help in making it easier to start playing with these things without gluing everything in place. O-like happens to have the crystals, though they're actually 5x5mm, not 3x3mm. The CD is likely to be 3x3mm or 3mmØ, but the actual dimensions are 5x5mm, and the HR face is unmarked (yet another reason not to glue things).

(Point being... actually getting started is the fast track to green, and some bits and pieces are helpful in that regard.)

Nice article, by the way.

I was hoping to one day build a high power green unit, I keep seeing crystals on ebay, I even saw one that was mounted in a commercial frequency doubler (Spectra Physics??)

Does anyone know if these crystals will double other frequencies? Or do they "resonate" at only specific frequencies?

It depends on the doubling crystal. There are various types and some have a broader range than others. Some are heavily temperature dependent and will produce usable output over a very narrow range of temperatures. For green lasers the most common doubler is KTP. It's not too temperature dependent compared to some others and it's relatively efficient, but it's efficiency drops off below 500nm or so..

That answers a lot, Thank you.

Explains why multiple wavelengths are not offered with KTP.

Ben,

I read it carefully,but I don't thikt I catched you meaning completely.
Send me some drawings about the cavity layouts. could you please.

Besides,I am trying to build a laser for cutting through Cu and Cu alloy.
the metal thickness range 0.1mm -1.0mm.
What's you comments. UV 355 / UV 266 or ????

Read this: Sam's Laser FAQ - Solid State Lasers

Everything you could want to know is in there, including drawings of cavity designs..

Enjoy!

Copper is a tricky metal to cut. There is very good absorption around the 700-900nm region, but very little absorption in the 500nm and below range. It seems as though the best option may be either a fiber laser or Alexandrite laser.

Thanks for the sticky.....

I want to clarify one thing that I have seen mentioned many times.

Frequency doubler units made by Coherent, Continuum, or Spectra Physics, such as this:
Newport Spectra Physics Frequency Doubler Model 3225-1 - eBay (item 250553771290 end time Jan-01-10 21:04:46 PST)

WILL NOT MAKE 532nm LIGHT IF YOU SHINE A POINTER OR HIGH POWER 1064NM LASER INTO THEM!

These doublers are single pass meaning they have very low efficiency at low powers. They are designed to double the output from high energy q-switched lasers only.

tnx for that beautiful idea!!!

Excellent writeup +1

yap!!! can u pls. teach me how to build

Ok, question. I bought a green laser pointer and have been thinking of seeing what can be done with it! I have a plan, there's a 808nm 1W diode on sale cheap enough and I'm thinking of seeing if it can take it. Firstly, the heat, a problem I'm already aware of but there aint much I can do about it, means it will have to be used in short pulses, as for driver board mods ECT I'm open to suggestions. Thing is, I'm hoping it can still run on 2AAAs

There's more to it than just slapping a bigger IR diode on on a 5mw pointer. Higher powers need larger crystals as well as definitely benefiting from proper heating and cooling.

Well, thing is this was more of an experiment than a proper project so I have slapped another 5.6mm can in with more power and it seems to have worked, trouble is it seems to perform best at room temprature and taking as long as 1 minute to heat up to full power from cold start. I put it in a flashlight housing and even though it looks kinda a crude effort, it still looks sweet
I estimate it puts out around 100mW or maybe more and with the focus set properly (Pinpoint) it can light a match, nothin really special but not bad considering the whole project cost less than $10!

Gonna post a guick how-to on it as soon as I can blag my girlfriend's camera to take a few shots

I think 80mw or so will light a match focused. But still an improvement over 5mw I guess. Be interested to see how it's lifespan is now though. Cheap lasers don't usually use the 'finest quality' materials, after all. I always wondered if you pulled one of the higher powered ones apart (the 100 or 150mw modules) heated the KTP, and cooled the YVO4 to the optimum temperatures and used a larger pump, how much power you could really extract from those crystals.