Laser. How much in this word ... And so on. I remember with what interest I opened one of the school physics textbooks and looked at the pictures of the ruby laser device. To do this would be akin to gaining the power of Engineer Garin's hyperboloid. How simple it was in the picture of the textbook! But to repeat this to a schoolboy in the 90s would be something from the realm of fantasy. Many years have passed, the Department of Quantum Electronics at LETI graduated, but the dream remained. It's time to implement it! So let's go.
As many people know, lasers are gas, solid-state, semiconductor, liquid, free-electron, gas-dynamic and, probably, some other. Personally, I have always been interested in a solid-state laser - a huge pulsed power and relative simplicity of design.
What are the components of a solid-state laser? First, we need an active element.
Most often, active elements are made from crystals of synthetic ruby, yttrium aluminum garnet (YAG) and yttrium aluminum perovskite (YAP), activated with neodymium, as well as neodymium glass (you may come across something else, but this is unlikely ).
Such active elements (hereinafter - AE) can be purchased (this is what was not available to a schoolchild in the 90s!) On avito, ebay, meshok or on specialized laser forums such as lasers.org.ru or laserforum.ru. It should be borne in mind that prices on laser forums are much lower than at flea markets and very often sellers there know exactly what they are selling. At flea markets, one in two sells "ruby" at surprisingly high prices, while they are not embarrassed by the slightly purple color of the "ruby" sold. Therefore, the first item when buying an AE will be its identification.
How to distinguish crystals from glasses? Usually, crystalline AEs have a smooth rod surface (there are exceptions: for example, for YAG and YAP, sometimes the surface is made with corrugations to avoid spurious generation) and do not have thickenings at the edges. Glass AEs, on the other hand, have a rough surface and thickened edges. There may be exceptions to these rules, but I don't know them.
Rubies often have uncolored areas from the ends of the rod - this is done because the ruby absorbs its own radiation, and since the ends of the AE will be in the crystal holder and the pump lamp will not get there, which for a solid colored AE will lead to the impossibility or a strong decrease in the efficiency of the laser ... Garnets, perovskite and glass with neodymium do not have such problems - they absorb their own radiation very weakly.
Some AEs can have small bevels that interfere with parasitic generation (or, with large bevels at the Brewster angle, give linearly polarized radiation). Enlightenment on the ends is also possible. I do not advise taking AE with bevels - it will be more difficult to align them, and in general, as far as I know, these rods are usually from amplifiers, and not from generators. There are also rubies with already applied mirrors on the ends. Such rubies were used in laser rangefinders and the purchase of such an AE will save you from searching for mirrors for a ruby and aligning the resonator, although the durability of such a laser will not be particularly great.
You should be warned against buying a large AE, especially a ruby one. It is very difficult to pump such AEs. Pomegranates and perovskite are the easiest to swing, they are heavier than glass, and the ruby generally eats the pump as if not into itself.
YAG
Ruby
Glass with neodymium
When choosing AE, it should be borne in mind that glasses with neodymium are silicate and phosphate (there are also a bunch of types of glasses, but I strongly doubt that you will find them on sale). Phosphates are more effective, but have lower mechanical and thermal strength. In general, in terms of thermal and mechanical parameters, any glass is much inferior to ruby and garnet and perovskite. The brands of glasses available to home-builders are LFS (laser (or luminescent?) Phosphate glass), LGS (laser (or luminescent?) Generating glass), KGSS (I think this stands for some kind of quantum-generating (silicate? Phosphate?) ) glass) or GLS (generating luminescent glass). LGS and KGSS are the old names for glasses. The most common glass is GLS-1 (it also corresponds to some number of the old KGSS and, possibly, LGS). Everything is fine with himbut unlike other types of glass, it is afraid of ultraviolet radiation. However, they are afraid of ultraviolet radiation and grenades with perovskite and even ruby. From it, they lose their efficiency, since the impurities always present in them are restored from ultraviolet radiation.
Also, garnet is slightly less effective than perovskite. You can distinguish garnet from perovskite using the fact that perovskite is polarized, which means that by looking at the picture on the LCD monitor through the end of the AE and rotating the crystal, you will see a change in light transmission from maximum to minimum and back. Pomegranate does not possess this property.
Perovskite and ruby give off polarized laser light (usually cut to produce polarized light). Garnet emits unpolarized radiation.
By the way, if you take a green laser and shine it into a ruby, it will glow bright red. Pomegranate and glass with neodymium emit in IR and you will not see their glow, although they will absorb the green laser beam as expected.
If you are wondering why glasses are so fond of with all their thermal and mechanical problems, then the whole point is in the possible size of the AE and the high concentration of neodymium, which is impossible in crystals - the activator ions have different sizes than the ions of the main elements of the crystal. Glass is an amorphous material and allows you to pump neodymium as much as you like. In addition, a higher lasing threshold is achieved in glass than in garnet and perovskite, which means that the laser will accumulate more energy before emitting.
So, we finished with identification. Now we need to figure out what we can hope for when choosing an AE. For ruby, you can get the generation of a bright red line 694 nm in a pulsed mode (in continuous output you will definitely not get it), glasses with neodymium work strictly in a pulsed mode (otherwise they are destroyed) in IR at 1062 nm for silicate glass and 1054 nm for phosphate glass, garnet and perovskite can be launched both in pulsed and continuous modes (depending on the amount of neodymium stuffed in them) in the same IR at 1064 nm. Neodymium also has other generation lines, but the main ones are those indicated above. Also, the gain of garnet and perovskite is an order of magnitude higher than that of glass. The generation threshold is also significantly lower. In ruby, due to the three-level pumping scheme, the lasing threshold is very high. At all,the lasing threshold is strongly related to the concentration of the dopant in the AE (neodymium for glasses and garnets / perovskite and chromium for ruby), the dimensions of the AE, and the parameters of the resonator mirrors. At the end of the article, I will provide a link to my program for calculating the threshold pump energy. Energy above the threshold will go into the laser beam with an efficiency of about 1%.
AE requires a resonator. The easiest way is to assemble a Fabry-Perot resonator. It simply consists of two parallel mirrors. Mirrors, however, are needed not simple, but dielectric. One with almost 100% reflection, and the second with the required transmission (usually 50% for glass and 90% for garnet and perovskite in pulsed mode and 15% in continuous mode). For neodymium AEs, these mirrors at 1064 nm are littered with all aliexpress and they are quite cheap, just do not confuse transmission with reflection. Actually, for the pulsed mode, the mirrors must have high radiation resistance, but the Chinese are unlikely to tell you the parameters of their mirrors.
A surprise awaits you for a ruby. Despite the fact that the ruby laser was historically the first, it turned out to be inconvenient due to the three-level pumping scheme, and therefore few such lasers were made. You cannot buy 694 nm mirrors on aliexpress, but at flea markets the price will not please you (ten thousand rubles or more for a mirror). Nevertheless, such mirrors from GOR-100 (optical generator based on ruby, 100 J), albeit slightly scratched, were presented to me on one of the laser forums, for which I am immensely grateful to this generous person (his nickname is Silverray). There is, of course, an option to use mirrors from the carriages of a DVD drive (green in the light as the output and blue in the light as deaf), but I was unable to launch a ruby laser with them, although there is information about the success of such a solution.Historically, in a ruby laser, mirrors were simply coated with silver on the ends, but only chemists can do this at home. In addition, silver absorbs radiation and burns out, and its reflectance cannot be compared with the reflectance of a dielectric mirror.
Deaf mirror from GOR-100.
Laser resonator.
The resonator mirrors need to be tuned. This requires progress. I made some homemade moves, but I recommend buying ready-made ones on aliexpress (there they are for a CO2 laser) or at flea markets. The thing is that the screw pairs there are ground, not cut. The polished pair does not dangle and does not play. You will definitely not find this in construction products.
Industrial mirror slide Homemade mirror
slide
Lamp, reflector, AE are collected in one unit, called quantron. You can buy a Kvantron ready-made (for example, K-107, K-301), or you can make it yourself. The laser head reflector (and the laser head itself) can be made, for example, from ceramic fuse cases (the author of this idea, as I understand it, is Nerv from lasers.org.ru). Industrial reflectors also come in ceramic or mirrors. Mirrors burn over time. Ceramic does not burn. It is necessary to glue the ceramic cases of the self-made laser head together very carefully, since the glue that has settled inside the pair will instantly char when the pump lamp flashes. You might be tempted to take the lamp and the AE and just wrap them in foil. Yes, it's called tight packing and it works great! But the foil needs a thick one - food will quickly become unusable and scatter in flakes at best,and at worst it will begin to melt and eat into the lamp bulb with dark spots.
.
-107 ,
A solid-state laser is usually pumped by a lamp or other laser. Our option is a lamp. Lamps are available for continuous and pulsed lasers. Domestic lamps for continuous pumping are marked with DNP (arc, for pumping, with a straight body) and are krypton lamps. I didn't work with them. Pulse pumping is carried out by xenon lamps of the ICP series (did not work with them), IFP (pulsed, photo-illuminating, with a straight glow body) and INP (pulsed, for pumping, with a straight glow body). For INP lamps, the diameter and length of the discharge gap are indicated. For example, INP3-7 / 80 has a length of 80 mm and a discharge channel diameter of 7 mm. The IFP series is marked according to the maximum energy, for example, the IFP-800 is a lamp for a discharge of 800 J.
I highly do not recommend looking at these lamps at the time of the flash!
For comparison, the discharge energy of the Soviet "Chaika" flashlight is only 25 J. And then 800 J. And then there is IFP-5000 ... and IFP-20,000. :) The required flash time of the lamp is usually in the region of 1-10 milliseconds. One can guess that the lamps get very hot during operation and, like the AE, they have to be cooled with distilled water. However, if you rarely give impulses, the lamp itself will have time to cool down. By the way, there is a lot of ultraviolet radiation in the spectrum of these lamps (note for disinfectants - in a millisecond any virus and bacteria will simply evaporate, however, often together with the surface - dark paper, for example, charred), which is harmful for crystals and GLS-1 glass, as I said above. This ultraviolet is cut off either by additives in the lamp cooling solution, or by a coating applied to the lamp balloon (like the INP3-7 / 80 A lamp), which, alas,tends to burn out without cooling. I don't use water cooling in my laser yet. Air is not worth using, since it requires good air purification, otherwise a speck of dust on the AE, mirrors, lamp will lead to burnout at its location. And you definitely do not need this. For ruby, the length of the discharge gap in the lamp must be not less than the length of the active (colored) part. For neodymium, a smaller length of the discharge gap is permissible.For neodymium, a smaller length of the discharge gap is permissible.For neodymium, a smaller length of the discharge gap is permissible.
Lamp INP3-7 / 80A.
To start the lamp, a battery of combat capacitors is required for a given voltage (depending on the lamp and is usually about a kilovolt and above) and an igniting pulse of ten or two kilovolts, which ensures the breakdown of the channel in the lamp. There are two ignition schemes: external ignition and sequential ignition. For external ignition, a high-voltage pulse is applied to a nickel electrode wound on the lamp bulb. For serial, an ignition transformer is connected to the lamp supply circuit. My choice is sequential ignition - there are no exposed electrodes outside the bulb. At this stage, it is worth thinking about wires that can withstand such voltages and currents. I chose PVMP-4 with a section of 0.75. The section is not enough, of course, but so far there is enough and besides, they can be connected in parallel.
External ignition
Sequential ignition
An ignition transformer can sometimes be bought (for example, the TIS-3 brand), but I first made it from the core of a fuel assembly, and then I assembled another new transformer from four glued ferrite rings with an outer diameter of 4.5 cm, an inner diameter of 3 cm and a height of 1.5 cm each, after gluing wrapped with fluoroplastic tape. On this frame, I wound a PV-1 wire with a cross section of 1 mm ^ 2 in the amount of 17 turns. All this is generously flooded with epoxy, because there still slips a dozen kilovolts. The primary winding is made of one turn of wire PV-1 with a cross section of 4 mm ^ 2 (in the photo below is an old transformer based on a core from a fuel assembly, in which the primary is just a bolt - it works worse than if you wind a full turn), on which the discharge capacitor 2 is switched μFx1500 V, creating an ignition pulse in the secondary of the transformer.This pulse can be used to instantly discharge the main battery of capacitors into the lamp, or they can first ignite a pilot arc in the lamp into which then drive energy for pumping at the required frequency. A device for such an arc is called a simmer, if the arc burns constantly, or a pendosimmer, if the arc does not burn constantly, but flares up some time before the discharge. The duty arc greatly increases the lamp life, but I have not done it yet. So my option is to discharge the capacitor bank immediately into an ignition pulse.and flashes some time before the discharge. The duty arc greatly increases the lamp life, but I have not done it yet. So my option is to discharge the capacitor bank immediately into an ignition pulse.and flashes some time before the discharge. The duty arc greatly increases the lamp life, but I have not done it yet. So my option is to discharge the capacitor bank immediately into an ignition pulse.
In general, my lamp power supply has the following circuit: Laser power supply circuit. ATTENTION! Do not install capacitor C1 !!! This is a common step-up push-pull converter from 25 V to 1600 V. During the tuning process, my thyristors often burned, and I experimentally found out that if you control the thyristors with a series of pulses, they almost do not burn out. “Almost”, because once in four months such an incident still happened and I added an additional protective chain after this incident. A choke in the lamp supply circuit is needed to ensure a "soft" discharge of the lamp. The calculation of the optimal parameters for the ignition of lamps is in Vakulenko's book "Power Sources of Lasers", and part of this calculation is in my program for calculating the threshold laser pumping energy.
: ! , ! - .
What energy will need to be stored in capacitors? Well, at least not below the threshold. My capacitors store at least 600-800 J. Peak - 2200 J. Capacitors, by the way, are very desirable low-inductive, which means that electrolytes for pumping are bad. However, the lifetime of the level in ruby is 3 ms, so electrolytic capacitors are good for a ruby laser, only they need to be shunted with an ordinary capacitor so that the backward wave when the lamp is ignited does not pass through the electrolytes and cause their degradation with a subsequent explosion inside the can. :) Yes, I had that. Therefore, I now start the laser with my headphones on - and so after the coronavirus the whistling in my ears / in my head has not completely gone away, and after the "bang" the whistle only intensifies and then requires treatment again.For electrolytic capacitors, you can also abandon the choke - they are already very inhibited.
For a neodymium laser, capacitors of the K75-40b type and the like are desirable, since the lifetime of the level in neodymium is less than a millisecond (the exact value is different in different media).
Photo of my power supply. The transformer is still old on TVS (it's just that I'm not at home, so I can't reshoot the pictures).
Do not forget also about safety glasses - the second eyes are not included in the set, as you know, and we don't have Tleilaxu masters. I bought POSOMZ ZN22-SZS22 LAZER 22203. For 694 nm, their optical density is 3 (weakened by a thousand times), and for 1064, optical density is 6 (weakened by a million times). Of course, looking directly into the beam with glasses is absolutely unacceptable!
Laser protection glasses
The base for the laser should be a massive slab. The more massive and durable it is, the better, because the alignment accuracy requires about 10 arc seconds for ruby and somewhat coarser for garnet, perovskite or glass. Garnet and perovskite generally forgive obliquely placed mirrors - their gain is very high and not so many beam passes in the resonator are needed. Even the output mirror for garnet and perovskite in the pulsed mode can be replaced with a simple glass plate (about 10% reflection from two faces). So I recommend pomegranate and perovskite to home-builders first of all! Will not disappoint.
Having collected everything on the base, the laser needs to be adjusted, that is, to set the mirrors parallel to each other and to the ends of the AE (this is important - oblique reflection from the ends will reduce the energy of the beam). How to do it? Take a regular laser pointer (this will be a pilot laser) and a mirror pancake from your hard drive. Glue a 1-inch plastic triangle to the pancake (cut the square into two parts diagonally and the part that you sawed and glue). Drill a hole about 0.5-1 mm in the center of the corner (and through the mirror pancake). Stick a pointer on the corner so that the beam passes through the hole. Hold the pointer with the glued pancake in a vice or put it on a tripod from the camera (here you have to be smart with the mount, but this option is much more convenient - you can easily change the angles and height). In this scheme, the pancake will play the role of a mirror, reflecting back to you the beam reflected from the mirrors,because at the end of the corridor it is very difficult to look for the beam with your eyes, but here it will be reflected almost next to you). And then turn on the resulting pilot laser and move it away from the adjusted laser by a couple of meters. Align the beam in height so that it passes through the resonator, and by adjusting the mirrors and turning the AE (together with the laser frame) on the ceiling, the reflections of the end of the AE and mirrors, driving their reflections on the hard drive plate to the exit point of the pilot laser beam. Aligned? Well, that's the whole alignment. Relatively rough, of course, but often works the first time. It will then be possible to adjust the prints by launching at least a slightly working laser. Problems can arise when the laser works near the threshold - there you will see a print on the target - the energy is low. But there's nothing you can do about it, you have to try to hit. Autocollimator, of course,it would be much more convenient and accurate to adjust, but where can you get it at home ...
And now the results
Ruby laser beam on a Ruby target
.
GLS-9P 12x260 glass with neodymium
Neodymium Garnet for Metal
Neodymium garnet on plastic
In accordance with the book "Lasers on yttrium aluminum garnet with neodymium" (Zverev, Golyaev, Shalaev, Shokin. 1985) I made a program for calculating the minimum threshold pump energy for neodymium and ruby lasers.
Program for calculating the minimum laser pump energy and lamp flash parameters.
Everything seems to work in the calculation, but the reliability of the calculation results is unknown (this is what I suggest you evaluate).
There are also features:
- I do not know the quantum yield of luminescence for garnet. I took it as 0.59.
- The duration of the flash pulse is counted as 2 * sqrt (L * C) and is not automatically transferred to the "Flash time, s" field. You need to do this with pens, if you agree with the time obtained in the calculation.
- I do not check the type of the entered data and their ranges. Perhaps I will do it later, if the program still gives reliable results.
- For glass with neodymium, I do not know the population of the lower laser level. I scale the population in a garnet with a known neodymium concentration to that in glass or garnet.
PS Some of the pictures in the article are taken from the Internet and belong to their authors.
PPS Thanks to the entire Lasers.org.ru forum for helping me put all these lasers together.