Its output is at 10.6 um (10,600 nm) or mid-IR. This wavelength is about 10 to 30 times longer than the other lasers under discussion and often considered a source of a heat beam than a light beam. At this wavelength, normal glass and plastic optics are either too lossy or totally opaque so alternatives must be found both for the end-mirrors and any other mirrors, lenses, or prisms required by the external optical setup. Divergence/diffraction effects are also increased by this same factor so obtaining a collimated beam is also more difficult.
Many common materials including wood, paper, plastics, composites, and properly prepared metal surfaces absorb quite well at this wavelength so the CO2 laser makes an effective marking, cutting, welding, and heat treating tool.
See the chapter: Carbon Dioxide Lasers for more information on the characteristics and applications of these devices.
It is possible for an amateur to construct a working axial flow CO2 laser (non-sealed, see the chapter: Carbon Dioxide Lasers) in the 10 to 50 W range without too much difficulty - at least compared to some of the other types of lasers described in this chapter. A vacuum system is needed but the range of operating vacuum is modest - 10 to 100 Torr. And while several gases are needed, the purity of the final gas fill isn't nearly as critical as for, say, the HeNe laser, and pre-mixed gas is readily available. See the section: More on Obtaining Gases.
With a bit of resourcefulness, no fancy glass work is needed. The power supply can be just a neon sign transformer on a Variac. The required mechanical precision isn't as great either so even if your machining skills are quite limited, adequate mirror mounts and structural components can be fabricated relatively easily. And, unlike the other traditional gas lasers (HeNe, Ar/Kr ion, HeHg, CuCl/CuBr, and the like), once constructed and aligned, the CO2 laser requires minimal maintenance and can potentially be a useful tool for real applications IFF it is packaged appropriately and provided with essential safety interlocks and protection devices. In fact, some commercial axial flow CO2 lasers are just refined versions of what an amateur can build. See the section: Descriptions of Typical Small Flowing Gas CO2 Lasers.
My only complaint about CO2 lasers in general are that the beam is totally invisible and boring in some ways. :-) Otherwise, it is nearly the perfect choice for a home-built laser - high power, continuous operation, and relative simplicity!
Although, constructing a CO2 laser is easy relative to other types of lasers, it is still NOT easy in an absolute sense. You will have to make a very considerable commitment of time and effort including dealing with the frustrations when, as is inevitable, some things don't quite go according to the script! The costs in $$$ can quickly become significant as well if you don't already have any of the supporting equipment (e.g., vacuum pump, power supply components) and aren't a first rate scrounger. :)
If all you want is a reliable working laser, save your money and buy a used CO2 laser system. Something similar to what you will be building can probably be acquired for under $1,000, perhaps less depending on your resourcefulness and luck. For example, see the section: How to Get a Laser Without Really Trying - Part 1. If you can afford a bit more, try to find a sealed tube CO2 laser system. These have essentially zero maintenance, and no requirements for a gas supply or vacuum system. New systems in the 25 W range can be had for around $5,000 from companies like Parallax Technologies. This is certainly not cheap but may be in the ball park once all your costs and time are factored into the equation. However, note that in either case, there could still be considerable costs associated with the beam delivery system for your application. For example, focusing lenses to get power to that wood stock can easily run an additional several hundred dollars.
You may easily spend hundreds of hours in constructing a working CO2 laser - especially if it is your first home-built laser. Note that while this estimate may seem high, it doesn't only include the actual fabrication of the laser head itself (tube, mirror mounts, and frame) but the effort required to select, acquire, and set up the vacuum system, gas supply and metering system, power supply and possible added branch circuit(s) to handle the electrical load, cooling system, and support structure (stable table or frame). And, providing and preparing the physical space: The home-built CO2 laser is not likely to be in any way portable and the kitchen table does not make a good work area. :) Once everything is put together, all the leaks are plugged, the parts that got trashed in shipping are replaced, you will have to align the mirrors (best done with a helium-neon laser - another expensive if you don't own one already), start feeding gases, apply power, and hope everything holds together!
Even after you achieve 'first light' (or I suppose, more correctly, 'first IR'), you will then spend countless more hours in getting the laser to be a reliable system - and it is quite conceivable that this goal may forever be elusive. If your heart isn't in the laser construction aspects of this endeavor, you will get discouraged very quickly.
If you DO decide to build your own CO2 laser, don't shoot for the stars, at least not at first. Begin with a modest size tube of say, 2 feet in length, which should still be good for perhaps 20 or 30 watts. There won't be any problems powering it with a 12 to 15 kV neon sign transformer and it will be small enough to easily fabricate - and store! Once you have gotten that operational, you will have solved all the tricky problems of vacuum, gas supply, mirror mounts, and cooling. It is easy to scale up something that works. However, starting with grandiose plans that fizzle will just lead to a box full of parts and equipment gathering dust in the attic.... :-(
(From: Flavio Spedalieri (fspedalieri@nightlase.com.au).)
Just to keep the expectations down, and also re-capping on the reason why we are building lasers in the first place.
For those who are building lasers for use as a tool, e.g., for metal or wood cutting, please be aware that you will be introducing a number of factors that will make it difficult for the final application.
Lasers for cutting applications should be producing a considerable amounts of power (around 100 Watts and up) to achieve a nice clean cut. At lower powers, it will take longer for the laser to cut the material, either if thin metal and or wood, and will cause burning and/or bad cuts. Another reason for having output powers in this range, is that at this wavelength, materials will behave differently (in terms of light behaviour) certain materials will reflect the IR energy, thus you are loosing energy through partial reflections.
Cutting lasers generally have to be very rouged, so to be resistant to fibrations that can cause mis-alignment of the optics.
Commercial CO2 lasers, are built so to be quite resistant to environmental changes. Also, many of the controls like vacuum, voltage/current, gas pressures/mixtures are electronically controlled.
For cutting applications, the laser beam must also be directed to the work piece, and this is done through an articulated arm - an optical waveguide. At the end of the arm, is the objective lens. These optics require to be enhanced for higher power IR beams, also it's not very easy at all, nor that trivial to build an articulated arm from scratch - much alignment involved.
Please be aware, that home built lasers are much more sensitive to vibrations, and are not of the 'plug-in and leave' variety. Home built lasers require much TLC, adjustments and continual tweeking, plus there is other things that have to be taken into consideration like: gas supply, mixture and gas pressure; power supply; mirror/tube adjustments and mounts; and thermal expansion and stability of the overall laser assembly.
Take precautions to avoid exposure to the direct or reflected beam. Since it is invisible, this means clearly labeling where it is and enclosing these in materials opaque to 10.6 um IR (these include glass and most plastics so it isn't that hard and doesn't obscure the action. Just because you can't see it, don't assume there is nothing there!
If you are using water cooling - either tap water or a closed loop system, make sure that water circulation paths are well insulated from the high voltage and that all fittings are securely grounded. Tap water is a fairly good conductor of electricity - think of it as a very soft wire. :) Unless you have a totally closed system filled with 100 percent distilled de-ionized water, there can be enough current flow to be lethal. Water and electricity do not mix!
(From: Chris Chagaris (pyro@grolen.com).)
Zinc selenide is an extremely toxic substance by ingestion and/or inhalation of dust, with a cumulative effect. This of course could be brought about if the optic were cracked or damaged in one's presence. This material liberates highly toxic hydrogen selenide in contact with gastric juices. Because of the material's reactivity with acids, it should NEVER be cleaned with even a weak acid, as these highly toxic H2Se vapors may be formed. Dermatitis may result from prolonged contact - don't play with your output couplers. :-) Although this material is practically insoluble in water, the use of gloves while handling the material is recommended. Gloves or finger cots should naturally be worn when handling any precision optic, as mentioned above. Particular care must be exercised when machining and creating dust or particles, not that I would think anyone will be machining their precious output coupler. :-) Symptoms of ingestion include garlic odor on breath. Ye best friend should tell thee. :-)
For more information, see the chapter: Laser Safety and the more specific information in the section: CO2 Laser Safety. Sample safety labels which can be edited for this laser can be found in the section: Laser Safety Labels and Signs.
(There may also be another book by the same author called just "CO2 Lasers" with more design information or it may be the same book.)
(From: Michael Andrus (andrus@ccountry.net).)
The summary below is for the CO2 laser from "Light and its Uses". Also see the section: Iannini CO2 Laser Description.
Refer to Typical Home-Built Carbon Dioxide Laser Assembly for a simplified diagram of the overall glasswork and power supply electronics.
CO2 can be easily obtained from dry ice; N2 from filtered air (despite the presence of O2 and other trace gases) apparently works well enough.
Unless you have your heart set on experimenting with various gas ratios or other gases, the use of commercial CO2 laser mix really is the best way to go as it eliminates the hassle and expense of a fancy gas mixing/metering manifold and the three separate sources of gases. CO2 laser mix is readily available from gas suppliers. See the section: Sources of Special Parts and Supplies for the Home-Built CO2 Laser.
Estimated gas consumption is roughly .1 liter/hour of helium per watt of output power. The amount of the N2 and CO2 then scales with their fill ratio. All of these flow rates are referenced to standard temperature and pressure (20 °C, 1 atmosphere). So, a 50 W CO2 laser will require roughly 0.625 liter/hour of CO2, 1.25 l/hr of N2, and 5.0 l/hr of He assuming the 1:2:8 gas fill ratios - or about 6.875 l/hr of CO2 laser mix. These gas flow rates don't sound like much until you realize that they will be almost 200 times greater at 4 Torr but of course the gas usage is still that relatively low number! :)
The idea isn't to pump CO2 laser mix through the tube fast as possible but to maintain a controllable pressure inside the tube with a slow flow rate. The gas doesn't get "used up" that quickly by the laser. Think of the pressure reguator and vacuum pump as an ideal voltage sources with the gas pressure regulator, tube, and metering valves as resistors. There should be a constant flow (current) through the tube and nearly constant voltage (pressure) across the tube. Thus, the resistance of the tube should be low compared with that of the rest of the system.
Pressure Supply Laser Vacuum Vacuum
Gas Supply => Regulator => Metering => Tube => Metering => Pump ==> Exhaust
Valve Valve
V1 o---/\/\-------/\/\------/\/\---o V2
R1 R2 R3
D1
H o--------+ T1 T2 +--------+--|>|--+---o HV+
)|| ||=||( | |
Variac )<--------------+ || ||( | D2 |
0-115V )|| )|| ||( +--|--|>|--+
15A )|| Neon Sign )|| ||( | |
)|| Transformer )|| || +--+ | |
+--+ 12kV,100mA )|| ||( | | | D3
| )|| ||( | | +--|<|--+
N o-----+-------------------+ || ||( | | |
||=||( | | D4 |
| | +--|--+-----|<|--+---o HV-
| | |
G o---------------------------+---+----+
_|_
////
The laser can be operated on half-wave rectified DC with about half the output power making for a very simple power supply. It can also be operated on the raw output of the neon sign transformer (AC) but the efficiency and maximum power output may be lower (though it isn't entirely clear why this should be the case since the discharge current is varying in the same manner for both cases). Even a low current oil burner ignition transformer (typically 8 to 10 kV at 15 to 20 mA) or smaller neon sign transformer will be adequate for testing of shorter tubes.
With the addition of some filtering and a ballast resistance to a basic rectified DC power supply, the nearly optimal current could be maintained continuously boosting efficiency and power output. To achieve a stable discharge, a ballast resistor of 100K or more will be needed. Based on its resistance and the tube current, the power disspation will likely need to be 100 W or more. A series/parallel combination of lower value lower wattage resistors will be the best option for its construction. A more sophisticated alternative to the huge ballast resistor is a regulator but doing this in the HV output would present some significant challenges, to say the least. :) Most modern CO2 laser DC power supplies are switchmode high frequency inverters using pulse width modulation to implement the regulation circuitry.
Once a filtered DC supply is used, a starter like a high power version of that used for modern HeNe laser tubes will be required. See the HeNe power supply information in the sections beginning with: Starters - Voltage Multiplier, Pulse, Inverter, Piezo.
Also, see the sections starting with: Power Supply Components for additional comments relating to the electrical characteristics of CO2 lasers.
Many variations are possible to this basic design beyond improving the power supply. These include changes to the resonator diameter and length; the addition and/or substitution of other gases, ratios, and pressures; and better optics. With the relatively non-critical vacuum requirements ease of mirror alignment, making these sorts of modifications should be fairly painless.
However, don't go overboard - at least not on your first attempt. Keep it relatively small, simple, and get it to work! Once you have a functioning CO2 laser, it will be relatively easy to scale it up. However, as the tube length is increased, the power supply voltage may need to be boosted as well. A 15 kV neon sign transformer may not be able to power a 1 meter discharge tube (around 0.8 meter appears to be the limit) reliably over the pressure range required for CO2 laser operation. If you are really interested in higher power lasers, consider a design using a split discharge tube. See the section: Steve's Notes on Boosting CO2 Laser Power and Quality Construction.
(From: Steve Hardy (hardy@sweng.stortek.com).)
"CO2 lasers operate at low temperature, so even soda glass could be used. Most people opt for borosilicate, since it's quite cheap and has better heat resistance (less thermal expansion). A quick back-of-the-envelope calculation shows that the temperature on the inside of a typical axial flow CO2 laser is only 20 °C above the cooling water temperature. Nevertheless, there may be hot spots (e.g., at the ends) which may be enough to crack soda glass.Regarding other materials (notably plastic), these could be used but have the disadvantage of lower thermal conductivity, lower stiffness, and higher price than glass. Fused quartz, alumina or beryllium oxide could be used, but the performance gain would not be worth the enormous expense."
Many people ask about where to get the bellows used to join the mirror mounts to the discharge tube. While, these are available from various industrial supply and optics companies, they will be expensive. However, there are a variety of low-tech alternatives including flexible water or natural gas line - readily available and low cost at your corner hardware store or home center. To be reasonably flexible, these may have to be longer than a real bellows, but then your mirrors will be further away from the bulk of the discharge and less likely to suffer from sputtering damage!
(From: Michael Andrus (andrus@ccountry.net).)
"For the bellows, I used pieces of flexable metal gas line. You can get this at a hardware store for a few dollars as opossed to $40 each for real bellows. This avoids the alignment problems with an O-ring-only design."
Try Janos Technologies for inexpensive but good CO2 optics and professional design help. Other sources for CO2 laser optics include II-VI Incorporated and Design Research Optics.
For external (e.g., beam steering) mirrors, I've even heard that the platters from harddrives work well, at least at low to medium power. While these are usually very slightly textured to minimize stiction, since the CO2 laser wavelength is over 10 times longer than visible, this is likely not a serious issue and doesn't produce much scatter. However, this should be confirmed for the specific platters being used.
(From: Flavio Spedalieri (fspedalieri@nightlase.com.au).)
I purchased my mirrors from Design Research Optics, 1-401-885-7353, Contact: Bruce Sunderland:
Part # Description Price
---------------------------------------------------------------------------
2-Z-80 Zinc Selenide output coupler $100.00
1.000" Dia x 0.120" Thick
TRCC-1012-S-10-ES Silicon Total Reflector $ 80.00
Enhanced Silver, 1.000" Diam.
x 0.120" Thick, 10 m radius.
Meredith Instruments current (March, 1999) has some CO2 optics at very reasonable prices ($49 each for ZnSe HR and OC, ZnSe lens, and silicon mirror). Check out their Miscellaneous Laser Parts page.
(From: Sam.)
Here's another company that lists CO2 laser optics but they probably aren't cheap: Laser Research Optics.
The following may also be useful though I don't know if he has anything other than lenses (e.g., mirrors) or smaller quantity pricing but it won't hurt to ask:
(From: Mubashir Khan (MubashirAK@aol.com).)
I would like to offer my services for such people who need inexpensive ZnSe optics. My manufacturer can produce such components in small quantities also. This offer is primarily for amateurs and hobbiests but others can also benefit. As an example, I have ZnSe plano-convex and plano-concave lenses starting at US$55/piece for 5 pieces. For more detail I can be contacted via email. I am with MnA Optiks, Postfach 1224, 64560 Riedstadt, Germany. Phone: +49 6158 915696, Fax: +49 40 3603 763573.
(From: 3jl (jjjimenez@terra.es).)
The construction of a mirror for a laser is not an easy work.
So, commercial mirrors are made of gold with several layers of ZnSe and NaF (or BaF) via sputtering, and have a reflectance greater than 99.5%. The problem is the cost.
I have been able to solve those problems buying the mirrors in Ebay. Make a search with "CO2 LASER". You could get a surprise.
For operation on DC, a bridge or full wave rectifier can be constructed from microwave oven HV diodes (typically 12 to 15 KPRV, .5 A). Depending on the output of your transformer, either 1 or 2 will be required in series for each rectifier (1.414 * VRMS minimum PRV). If you get them from the same lot number, equalizing components are probably not needed where 2 are in series. Electronics service parts places like MCM Electronics and Dalbani usually sell replacement diodes for about $2 each. If course if you want to construct your own, a stack of 1N4007s will be even cheaper. See the section: Standard and Custom HV Rectifiers for more information.
In addition, if you want to drive your tube with filtered DC, smoothing capacitors as well as high power ballast resistors will be needed. Electronics distributors or electronics surplus outfits are the most likely source for these. Of course, you could get really fancy and add a current regulator. There is some discussion of this in the chapter: Carbon Dioxide Lasers and some circuits that may be useful as a starting point can be found in the chapter: HeNe Laser Power Supply Design. However, these are for the advanced course. :)
Note that if the coolant comes in contact with the electrodes, it needs to be non-conductive. Distilled water may be acceptable. Mineral oil is an excellent insulator (if kept dry) as well and often used for cooling transformers and other electrical equipment but is flammable. However, it isn't perfect:
(From: John De Armond (johngd@bellsouth.net).)
Disadvantages of mineral oil: Lower heat capacity than water, a bit harder to pump, flammable, softens some polymers, it is conductive when contaminated with moisture, may polymerize under the influence of UV from the discharge tube, especially if there are any paraffinic fractions in the oil, doesn't magically dry up when you kick a can of it over. :-) I'd certainly use transformer oil if I went that route but you still need to keep it dry.
As for pumps and such, check out C&H Sales (one of my favorite places), and other surplus sources.
For a circulating pump, it's hard to beat a carbonator pump. Around here, the used restaurant equipment places practically give them away. Coke will put fountains in restaurants and then when the restaurant goes out of business, they'll either not get the fountain at all or else they'll just get the head and leave the rest. In any event, carbonators are running out the yingyang around here. They usually come complete with a 1 gal stainless steel tank.
The little submersible pumps lack the head pressure you need to force a meaningful flow through tubing. These things are typically rated for only 20 feet of head or so and that is only a few PSI. The pump will churn and froth and look like it's working but it's not. you really need a positive displacement pump. If you can't find a carbonator pump, another option is the peristaltic pump. This is the familiar "roller and rubber tubing" metering pump. These things are cheap enough to be throw-aways. Indeed, I've had several different vendors in my restaurant for grease cutting enzyme drain treatment. Each vendor installs a peristaltic pump to meter the enzyme. And when they let the reservoir get empty and I fire 'em for not servicing it, they never bother to come get the pump. I have several on the shelf. What's really nice about these pumps is tubing can be had for almost any media. Silicone rubber is typically used but if you want to pump something really gnarly, any tubing that is reasonably flexible will work.
(From: Steve Roberts (osteven@akrobiz.com).)
The 25 to 50 W Sharplan lasers use a 22 litre per minute plastic pump made in Germany by a company called Eheim (model 1060). It's got a nonconductive case and rotor but is not self priming. They have a 1 liter reservoir above the pump to keep a constant level in the system. The radiator used is a 6" by 8" unit with 3 layers of fins, 1/4" tubing in the core, and a 200 cfm fan.
(From: John De Armond (johngd@bellsouth.net).)
Check the restaurant supply outfits near you. Just like the carbonator pumps, CO2 regulators regularly get tossed when the old fountain gets yanked. A couple of years ago I walked in one of my favorite used restaurant equipment places and found a large box of regulators labeled to go to the brass scrap yard. I bought half a dozen for their scrap value. (No, don't want to sell any. :-( $5 is the maximum I'd pay for one. Change the CGA nipple and away we go.
For a fairly cheap ($20 to $35) flow meter, check out either your local welding equipment supplier or companies like OMEGA Engineering, Inc..
The use of premixed gas also doesn't allow you to easily experiment with changing gases or the partial pressures of each gas. The obvious way around this is to provide separate tanks for each possible gas and a mixing manifold. Needless to say, the costs, complexity, and difficulties in controlling and monitoring gas flow can quickly get way out of hand.
If you really don't want to invest in official CO2 laser mix or high pressure tanks, regulators, gauges, and metering valves, there are alternatives, at least to get your laser going initially: CO2 from a chemical reaction (e.g., vinegar and baking soda!), CO2 cartridges, or seltzer makers; nitrogen from the air, and helium from party balloon tanks. Depending on what you end up with, gas composition and purity may be somewhat hit or miss, but CO2 lasers are fairly forgiving. As long as you use decent optics that are fairly well matched to your laser tube length, there is an excellent chance of getting the laser to work, though probably at much less than optical performance:
However, some companies may charge a lot for a single tank of CO2 laser mix:
(From: Curt Graber (cgraber@fwi.com).)
Sad to say but most of the larger corporate gas supply companies want to have a bottle a month relationship min. and if you sound like you have never leased/purchased a tank previously they automatically go into unobtanium mode and pricing.
doing your own mix is a great opportunity to tune your resonator for it's tastes in gas mix but will not save you any money at all as you need a significantly bigger manifold set-up and precision meters to correctly (read repeatably) set-up the mix as the gas is depleted from each reservoir at different working pressures and volumes.
My suggestion is to find a second hand medical laser supplier and buy a full or mostly full tank of pre-mix that was a "removed from medically obsolete" flowing gas laser, there are a lot of them out on the market! and the bottles typically get thrown gently on the shelf to await another lifetime and or lifestyle such as your application.
If you don't have any second hand medical equipment suppliers in your area scour the internet for a find close to home as you will want to likely also see their other wares stripped from similar units.
(Portions from: John De Armond (johngd@bellsouth.net).)
Generating the CO2 chemically could be kind of nasty unless you came up with some sort of drying scheme. Not only may water vapor interfere with lasing, the coating on some optics or the base material itself may be damaged by excessive moisture.
Why not use the CO2 cartridges for pellet guns or seltzer water bottles? You can buy a seltzer water maker at the local kitchen notions store for under $10. It'll puncture a CO2 cartridge and bleed it into a pint metal bottle. If you vacuum the bottle beforehand you'll end up with clean CO2 at a reasonable pressure.
Alternatively, I only pay $10 for a 20 pound tank of liquified CO2 from Coke for the fountain in my restaurant. That's high compared to what I could get it for if I owned the tank. Coke charges no rental and only a $30 deposit. If you have a favorite restaurant, perhaps you could get the owner to sell you a tank of CO2. Another possibility is a party shop. Party shops around here sell pony kegs of beer (about 1/3 the size of a regular keg) and loan the tapper (with a deposit) that consists of the fittings to the keg, the dispensing head and a small CO2 tank (about 3 pounds, I'd guestimate). Even if you have to buy the keg of beer, life could be worse! :-)
As for cheap helium, around here, Wal-Mart sells a party kit that contains a 30 pound Freon cylinder filled with He and some balloons for about $20. OK, so your really don't need the balloons. :) Some of the party stores carry it too but they charge more. Since it's a standard Freon tank, it has a standard SAE 1/4" flare refrigeration fitting. You can attach either a refrigeration service hose or simply a 1/4" flared copper tube. The tank is pressurized to about 275 pounds so there's quite a bit of helium in it and the helium is pretty pure. I use those tanks in my neon shop for helium flush gas on my pumping manifold. This is about the cheapest helium one can get, particularly when you figure in the lack of a cylinder rental (or purchase).
(From: Steve Roberts (osteven@akrobiz.com).)
A few suggestions:
I just gave away 2 medical surplus tanks at 3/4ths full to a local artist for his laser carving experiments as I'm not a CO2 person. I paid $50 a tank for one and the other was on the laser I scrapped. When my artist friend asked asked me what the refill cost was, I called a medical laser friend and he said, out the door cost with new tank and full tank ownership in St. Louis from his local gas supplier was $140 and tax. Now what he has is a local more independent gas company that buys bulk and blends their own mixes, instead of a big national chain. I'd find such a place, they do exist, go see em in person, not on the phone, present a reasonable appearance and if they ask you what partial pressures you think you need, be able to answer them. I bet you score a tank cheap.
Usually laser gas is considered a high purity mix and they need to vacuum out the tank, weigh it, specially clean it, then mass spectrometer certify it, etc. For a really precise lab laser or sealed service, I could see it needing that kind of preparation. But what my buddy does is have the local place do the mix on a normal fairly clean tank by pressure and percentages. They have no problems and the laser works just fine.
So I'd either talk to the biomed engineers at your local hospital or find a used Sharplan tank from a medical reseller or find a more reasonable gas place.
However, if you're mixing your own gases, for a flowing gas laser, normal 99.999% tank gas is fine. The people who tank the gas won't certify it as 99.999% but that's what it is, and usually better. A few parts per million of impurities won't hurt you. But, sealed off lasers are a whole different matter.
Unless you have a monster of a CO2 or are in daily production, the unit just sips gas, especially if you add the solenoid that kicks off the gas after the regulator when the plasma is off.
The local 5 kW CO2 job shop burns one big size S tank every four to eight hours. If somebody does laser cutting in town, hit em up for a tank at their rate plus a little extra. If you have one near by, they probably will help you out once or twice, as the more they buy the cheaper it gets! But watch out for demurrage charges, which is when a tank sits too long at one place.
Setting up a blending station at 1 to 10 Torr final pressures is a real pain in the neck, most needle valves, even the micrometer ones, are not that good, and I do high vacuum as part of my living. Just adjusting the final pressure needle valve on a flow gas is a dog. CO2s do lase over a broad pressure range. You can be both too low and too high and still get about 1/2 the rated power, but the sweet spot does vary some what with current.
I actually got some lasing once using a welding shielding argon/CO2 mix and blending in nitrogen, but 1.5 watts out of a 60 watt tube was kind of disappointing. :-)
(From: Tom Miller (tmiller@umaryland.edu).)
I picked up one of the He tanks from Wal-Mart today - $15.19 including tax. This tank is the size of the 15 pound freon tanks usually found in automotive shops. I think if we used these by adding a small amount of co2, only 4 to 5% more pressure, and bleeding air in for the N2/O2, it should make for a cheap setup.
(From: Sarlock T. Icedragon (sarlock@twcny.rr.com).)
I checked a local company out here and they will sell me a 20 pound CO2 tank, and the first fill with custom mix is on them (if the tank is bought from them of course). For this they ask a total of $117.
(From: John De Armond (johngd@bellsouth.net).)
That sounds like a pretty good price if he's actually selling you a new tank. Particularly if it is an aluminum one. (Luxor is the most popular brand of aluminum tank.)
Here are some additional comments on CO2 gas mix.
A gas filling station cannot legally fill a tank owned by a competitor. Many gas companies will "sell" (actually a one-time charge) you a tank. The tank will have their name on the collar and they take the empty and give you a full one each time. As is obvious, after the first filling, you no longer possess the same serial numbered tank as is listed on your "purchase" receipt. There is no problem until you take that tank and try to get it refilled at a competitor's station. When the competitor sees another company's neck ring on the tank and the tank serial number does not match the receipt you bring along, they typically refuse to fill it. Or if they do, they will not do a tank exchange and instead send your tank off for filling. That means you have to wait for it to be filled.
If you actually buy the tank, the neck ring will be blank. You can stamp your name on it if you wish. Your tank's serial number will match the one on your sales receipt. Your tank will be sent to to the specialty gas supplier of your choice, filled and returned.
All that is required to get any tank filled is (1) proof of ownership (sales receipt does just fine) and (2) current hydro (date is stamped on the cylinder). Since laser gas is a specialty gas, one won't be buying it from the local welding supply store so the issue of swapping an empty tank for a full one versus sending the customer-owned tank off for filling doesn't arise.
You don't need a fancy expensive regulator. A MIG gas regulator is less than $50 at Tractor Supply and other outfits that sell the Century welder and all of its private label versions. A helium balloon filler is even cheaper. CO2 laser gas is lo-tech, unlike some other types of lasers that require very pure gas.
As for shipping, if you have to UPS it, put it in a box, label it as compressed gas and ship. UPS charges a hazardous materials surcharge because it is high pressure (and because they can).
(From: Steve Hardy hardy@sweng.stortek.com).)
The typical ratio of 1:2:8 of CO2, N2 and He respectively was suggested by Linde Gas who supplied my mix. The precise ratio is 9:18:73, and is expressed in terms of partial pressure. In practice, this is fairly close to the molar ratio (since most gases at 1 atm and 20 deg C, are about 1 mole per 22.4 litres).
Please don't worry about precision. 10% deviation either way in any of the components is not going to make a noticeable difference - at least not to us amateurs.
You may hear of different ratios being used in various circumstances - usually the amount of He is varied. For example, a lower ratio of He is used in a TEA CO2 laser. You can even run your laser with 1:1 CO2:N2 and no He, with reduced efficiency.
(From: Sam.)
See Spectra Gases' CO2 Laser Mixtures Page for some typical percentages.
(From: Mark Dinsmore (tango@coolmail.net).)
I found what might be another source for CO2. We have a local paintball supplies store, and I bought a 12 ounce bottle of CO2 and valve, full, for about $40. 12 ounces of CO2 is a LOT of gas, and a refill is only $2.00. I saw the guy fill it from a standard full size siphon tank, so I believe that it is reasonably pure. I'm sure it's not research grade, but I'm hoping it's good enough.
Mirrors salvaged from commercial CO2 lasers will be satisfactory if their RoC and reflectivities are compatible with your tube length and they are in good condition. (CO2 mirrors are subject to both degradation during use due to high photon flux and from moisture/water damage while in storage.)
Although generally similar to the CO2 laser from "Light and its Uses" (see the section: Home-Built CO2 Laser Description, it is more than twice as long and should therefore be capable of much higher power operation. Some glass working is specified (though this could easily be avoided with minimal design changes) and the power supply has more frills (but is only AC instead of DC and therefore will not be as efficient).
This ultimate operating pressure of your CO2 laser depends somewhat on you final design. Long, small bore plasma tubes will effectively operate at higher pressures than shorter, fatter tubes. Long, relatively small bore CO2 laser tubes will operate at between 25 and 35 torr, at best.
The high pressure regulators normally attached to compressed gas cylinders will not offer the control needed to accurately 'feed' a CO2 laser. Needle valves must be incorporated in the lines after the regulators to adjust each separate gas flow to the laser head.
There are much simpler options available rather than using neon sign electrodes in this type of laser. Why do some people insist on neon sign electrodes when they are not really needed in a small slow-flow CO2 gas laser? If your laser is to be operated on DC then all you really need is a short length of copper or brass pipe in the plasma tube as a cathode and simply a small loop of wire as an anode.
There is a need for a massive electrode (copper pipe or similar) on the cathode-end of the tube, as electron/ion bombardment will tend to heat this end. The anode-end only requires a wire loop. I would leave the method of sealing the wire into the glass tube to those with glass-working skills, as there would be many ways to accomplish this.
As an experiment, I ran the CO2 laser using a 10,000 VDC, 30 mA power supply. Ambient temperature was 77 °F. The cooling water for the plasma tube started at 76 °F, as I had not kept my chiller plugged in to power over the past several weeks. After approximately 45 minutes of normal operation the cathode electrode (the one we're concerned with) was just warm to the touch (be sure to turn off power for that test!) and measured just a tad over 100 °F according to my 'not very accurate at that low temperature' temperature sensor. Don't worry about this if you are using a good-sized electrode.
A three electrode configuration does have some merits, but only if you have a special power supply designed to properly operate this type of design. You would ideally want the anode in the center of the tube and cathodes on each end. The cathodes must be large enough to dissipate the heat created by ion bombardment while the anode can be much smaller, as this does not heat up and is better suited to be in the center of the tube. This set-up would not be practical using a neon sign transformer, as the secondary high voltage windings are center tapped to case ground. More importantly, is that you would not gain anything using a single NST. You would merely be putting half of the output voltage on each half of the plasma tube. Water jacket construction also becomes much more difficult when dealing with a center electrode type configuration.
The easiest and least expensive method for attaching the electrodes to the CO2 plasma tube, which will require no glasswork, is to use the copper pipe fittings at both ends. This will also allow those who wish to operate on AC a simple solution. The common copper pipe fittings that are available at your corner hardware store will work well in this application. With a little modification these will also provide ports for vacuum and laser gas. If you attach alligator clips to your power supply output wires, these can be attached anywhere on the copper fittings for simplicity.
My copper fittings are attached to the plasma tube using an O-ring seal and a simple clamping method built from Plexiglas plates and machine screws. Two-part Epoxy is employed to attach the first Plexiglas plate to the outside end of the plasma tube and to attach the end of the copper fitting to the second Plexiglas plate. All the screws are placed 120 degrees apart. Mirrors are Epoxied onto the mirror adjustment plates and a thick O-ring will provide both a vacuum seal and enough play for mirror adjustment (no bellows required). This is a very simple and inexpensive method which works well and should allow anyone with a reasonable amount of mechanical skills to build a functioning CO2 laser without the need for exotic or expensive equipment.
I use a standard CO2 laser gas mix of 4.5% CO2, 13.5% N2, and 82% He. This is commonly available from commercial suppliers like Airgas, Inc which is where I obtained mine. These companies can usually provide whatever mix ratios you specify. Other mixtures that I have seen that will reportedly work well are: 8:8:84, 4:28:68, and 15:25:60 (CO2:N2:He respectively). I would guess that experimentation with varying ratios would lead you to the best mix to use in your particular laser.
My original CO2 laser and power supply is pretty much along the lines as the Scientific American as described in the section: Home-Built CO2 Laser Description
I have recently built a smaller and more compact power supply for it. This consists of a 9,000 volt, 30 mA neon sign transformer, manufactured by the Canadian firm of Allanson, as the basic unit. This is the type of transformer with the two high voltage terminals coming directly out the top of the tar insulation, all enclosed in a somewhat larger metal case. This case left plenty of room for a bridge rectifier, key switch, fuse, 'on' light, milliammeter and a 3 amp Variac. A six foot 3-wire appliance cord with IEC input module and two 20 inch high voltage output leads, with color-coded alligator clips complete the package.
This power supply seems to perform as well as the 15,000 volt @ 60 mA unit that I had employed in the past with this modest sized laser. I no longer have a supply of commercial CO2 laser gas and have been using the 'three separate gas' system for some time now with good success. I have been using a commercial cylinder of compressed helium (useful for most other lasers also), nitrogen from the air (bubbled through water) and sublimating dry ice for the CO2 supply. I had bought a small supply of dry ice on Friday which lasted only until Saturday evening, due to its tendency to mysteriously disappear even when stored in an insulated cooler. Out of sheer curiosity today I decided to try a much more readily available source of carbon dioxide gas to operate this laser with. Searching through the cupboards I found some white vinegar and a small box of baking soda. Lo and behold, a small amount of this mixture in a flask produced a fine source of CO2. I was pleasantly surprised to open the needle valve from this flask to the laser and see a fine burn pattern appear on my thermal FAX paper target. No more dry ice! I just thought you may wish to pass this on to the amateur laser community. This really simplifies the gas problem for small CO2 lasers.
All small CO2 lasers such as this operate under similar conditions. The gas mix is pretty much standard for most slow-flow CO2 lasers and is usually 4.5% CO2: 13.5% N2: 82% He. This should be available pre-mixed from any large industrial gas supplier. As far as optimum operating pressue, this should be easy to determine experimentally. It will depend somewhat on the plasma tube dimensions, but will likely be somewhere between 5 and 25 torr. If you have an adjustable output power supply, you simply start at a the lowest pressure and lowest current setting that will allow laser output, then slowly increase pressure as you increase current. You will find that power output of the laser will increase to a point of saturation. Just below this point, at the highest output level, is where you want to set your controls
Being able to see the glow discharge through a clear glass tube has advantages beside the purely aesthetic value. With some experience operating a CO2 laser, one can tell much about the operating conditions from the visual appearance of the glow discharge. The color, intensity, and density of the plasma all give clues to adjusting parameters such as gas mix, gas pressure and current density for optimum performance and power output.
I am not sure of the power output I am getting with this system, but it will easily ignite the FAX paper in a matter of five to ten seconds without any focusing optics. The burn pattern is quite large also with this wide bore (1 inch) plasma tube. I have read about your plan for measuring CO2 output power by heating a measured quantity of water for ten minutes. I have yet to try this yet, basically for lack of a suitable reflector to steer this beam into a container. I have a simple thermometer with a flat black penny attached to the stem to see temperature rise from the beam impinging upon said target. This allows some comparison of output power, but is there anyway to translate this to watts? I am getting about a one and one half °C rise in temperature per second at full power.
(From: Sam.)
Yes, by knowing the mass of the penny (about 5 grams I think) and the rate of initial rise in temperature, you should be able to calculate the approximate beam power. What you need is temperature rise/joule (or calorie) of energy for a standard penny. :-) I say 'initial' since losses due to convection and radiation of the heat energy will be small. This is still going to be tricky without a serious effort to insulate the 'sensor'.
Here are some notes on the construction and power output of my home-built CO2 laser:
Plasma tube:
I do not recommend having a plasma tube of 1,000 mm (or more) in length, there my be problems with sustaining the discharge. 15/30 NST will not be enough in striking a tube of this size - even a 15/60 may have some troubles.
I have cut my tube length from 1,000 mm to 800 mm.
For my tube, I chose a bore size of 20 mm ID and 24 mm OD. The larger bore will generate more power.
Optic/mirror adjustment mounts:
Refer to the following absolutely fabulous ASCII drawing when reading the description below and also see Flavio's Carbon Dioxide Laser Plans for some details.
_ _ _ _
HR | | | | Resonator Rod | | | | OC
Adjusting =[| |]=======[| |]===================[| |]=======[| |]= Adjusting
Screw _ | | | | (One of 3) | | Out | | _ Screw
O=| |=>| | ______|_|_____________________|_|__||__ | |<=| |=O
__|_/\/\_|__|_______________________________________|__|_/\/\_|__
| Bellows Plasma Tube Bellows |
HR|__ _ _ _______________________________________________________|OC
| \/\/ | |_ ____________________________________| | \/\/ |
O=|_|=>| | In || | | Water Jacket | | | |<=|_|=O
| | | | | | | |
=[| |]=======[| |]===================[| |]=======[| |]=
|_| |_| |_| |_|
My laser tube has an integral cooling jacket, and at both ends of the plasma
tube are 20 mm long extensions of the main plasma tube. These extensions then
go into two copper pipe fittings (one at each end. The solid fittings then
are soldered to a set of copper bellows, then another pipe fitting is soldered
to the other side of the bellow.
One end of the bellow assembly is connected to the plasma tube, and the other is soldered to a small brass plate. The brass plate then is screwed into a Plexiglas plate that forms part of the adjustment plate. Behind the adjustment plate (the plates closest to the center), are a set of fixed Plexiglas plates, these actually DO NOT support the plasma tube in anyway. In this plate, will be fastened a set of three fine threaded rods positioned in the typical 'L' configuration. The front adjustment plates are then spring loaded onto this plate - which enables the adjustment.
In the front adjustment plate, the front has been worked so that the optics sit in an 'O-Ring' groove, then a clamping plate is positioned over the optic to hold it in place - vac tight seal, and the ability to remove the optic and not mess up the alignment of the laser.
There are no requirements for specialized ceramic insulators required. And, best of all, the optics and their adjustments are isolated from the high voltage. But all caution must still be taken into account when working on the lase.
Estimated power output:
Although it is very difficult to determine due to the many variables involved, from my design using 1" optics, 20 mm bore, 800 mm length, and a power supply of 15 kV at 60 mA, I am guestimating an output from 20 watts up to 60 watts - but this is only a guestimate going on commercial lasers of similar size.
Providing more voltage/current will give you higher output, until you reach a saturation point where the output will begin to decrease with further input (the same as a HeNe laser).
(From: Steve.)
My laser was slightly unusual in that a filtered DC supply was used. This means I had the luxury of earthing the cathode and. If you use a neon sign transformer, you won't be able to do this - so please make allowances!
Tube Size:
I recommend the use of a smaller diameter tube. You can run them at higher pressure and higher current, since the smaller diameter provides better cooling of the gas. I used 15 mm ID tube, but only because of ignorance when building it. Nevertheless, I was very pleased with the results (about 30 W out for 600 W in).
Electrodes:
The electrodes do not get overly hot in operation. Most power is dissipated in the gas, which is largely conducted to the tube. The anode stays cool by normal convection. The cathode is subject to ion bombardment but this is not as bad as some people suppose, because the gas pressure is high enough that the mean free path is so short that ions don't have much chance to accelerate. Solder most definitely will not melt.
Aluminum is a good choice for electrodes. It naturally acquires an oxide film, which reduces the danger of sputtering. I examined my electrodes after operation and found them to be covered with an iridescent oxide film.
Vacuum stuff:
Laser gas is typically 1:2:8 ratio of CO2, N2 and He respectively. I have been assured by the experts at Linde that this is the only mixture you need to use for low flow lasers. This gas comes in steel cylinders, and has the trade name of 'lazpure' (at least for Linde in Australia).
The gas is compressed in a cylinder at many times atmospheric pressure. This has to be brought down to 25 Torr. You will need a regulator on the bottle. I happened to have an Oxy regulator which is just fine. The output of the regulator is adjustable between 1 and 4 atm. This is fed into the metering valve, which is basically an adjustable small orifice. I used a Porter metering valve (series 20) which costs under $100.
Metering valves don't necessarily cut off completely, so you need to be careful to shut the bottle off (using its valve) after use, to avoid damaging the tube or vacuum pump with overpressure.
The gas (quite cold due to its expansion from high pressure to a relative vacuum) flows rapidly down the tube to the anode end, where it is exhausted by the vacuum pump. Before exiting the tube, I have a slight restriction to slow the flow otherwise too much gas would be wasted. It would be handy to make this adjustable too (so you could control the pressure as well as the flow rate). If so, then make sure its gas conductivity is much, say 100 times, higher than the metering valve at the other end, because it is used at much lower pressure differential. One possibility which I haven't tried is to use a ball valve (normally used for water). Note that any parts which the operator needs to touch must be positively and securely earthed.
(From: John.)
There is a device available called the back-pressure regulator that makes regulating the pressure and flow much easier. A related device, the vacuum regulator will do the same job in this application. The back-pressure regulator works just the opposite of conventional regulators. It opens its port whatever amount is necessary to bleed off pressure in excess of the set-point pressure. With a back-pressure regulator between the vacuum pump and the tube, the tube pressure is with the regulator and only the flow rate is controlled by the inlet needle valve. This little valve will pay for itself in saved gas because with needle valves on both ports, one tends to juggle them until a stable pressure in the tube is achieved and then just live with the flow rate even if it is much too high.
C&H Sales usually has these devices surplus but they typically cost less than $100 new. About the cost of a tank of gas. :-)
(From: Steve.)
Sounds nifty. I presume the back-pressure regulator regulates the pressure differential across its inlet and outlet ports. Is there a type of regulator which can regulate to an absolute (but subatmospheric) pressure?
The reason I ask is that it would be nice to be able to set the tube pressure in absolute terms without worrying about the effect of the vacuum pump (since its inlet pressure would vary slightly with the flow rate).
(From: John.)
Two types are available. One type is referenced to atmosphere. The other has a vacuum capsule and is referenced to a vacuum and thus absolute pressure. Wallace & Tiernan (sp?) makes an absolute regulator but it is $$$$ new. I've seen 'em at C&H but not often.
(From: Steve.)
Note that with twin 'needle valves' I never had any trouble getting a stable pressure, but as you say it was hard to adjust flow rate and pressure independently. A regulator such as you mentioned would make life much easier.
(From: John.)
Yep. An alternative is the mass flow controller. A self-contained mass flowmeter and valve. Widely used in the semiconductor and thin film industries and so are fairly common on the surplus market. That way you set the flow with the controller and then set the pressure with the outlet valve.
(From: Steve.)
One problem would be actually measuring the flow rate. I thought of putting a balloon over the pump outlet, and measuring the increase in volume using Archimedes' principle etc. This would allow derivation of the flow rate in the tube by multiplying by the pressure ratio (atmospheric/tube). Or, just for fun, use a 'calibrated condom' and measure time-to-burst!
(From: John.)
:-) The common amateur scientist method is the soap bubble technique. This requires only a sidearm burette. A squeeze bulb is fitted to the burette and the gas flow is introduced to the side arm. Fill the bulb with a soapy water solution. When you're ready to measure the flow, simply squeeze the bulb until the water covers the side arm. A bubble will form in the burette. Simply time how long it takes to travel a set number of graduations. Assuming the graduations are in cubic centimeters, dividing the number of graduations transversed by the elapsed time in seconds gives the actual volumetric flow in cc/sec. This works best when the burette is vented to air. Assuming your vacuum system is fairly tight, you can flow the exhaust of your vacuum pump. This is especially practical if you're using a hermetic refrigeration compressor for the vacuum pump. Remember that the flow must be corrected to he tube pressure using the gas laws to get the flow through the tube.
A hot wire mass flowmeter is also very easy to make. Certain GM cars used a hot wire mass flow meter to measure intake air. If you find one of these in the junk yard, you'll see a small venturi containing a couple of hot wire filaments inside the flow path. Simply removing the integrated electronics from the housing and mounting the hot wires in a smaller tube is all that is necessary to make a very sensitive flowmeter. The integrated electronics output a pulse train whose frequency is proportional to flow. A computer interface is thus obvious. A cheap industrial pulse totalizer/rate meter can directly display the flow. The electronics require either 5 volts or raw 12 volts, depending on the model. Note that like all thermal instruments, the output is dependent not only on the flow but also on the conductivity of the gas so the instrument must be calibrated to the gas mix in use.
(From: Steve.)
The vacuum pump (and flow restrictor, if used) is connected to one end of a long tube. This tube must be longer than the discharge tube length to prevent arcing between the anode and the vacuum pump. The gas pressure in the connecting tube may be considerably lower than the discharge tube pressure, which tempts electricity to flow along it.
(From: John.)
Let me amplify on that a bit. If one plots voltage drop versus pressure, one finds that the minimum voltage drop, and therefore the longest possible discharge for a given voltage occurs at about 1.25 Torr. Below that it rises sharply until about 1 micron which is essentially a vacuum insulator. Above that point the voltage drop also rises but with a much more gradual slope.
To put this in perspective, a 15 kV transformer will run in excess of 100 feet of 25 mm tubing pumped to 5 Torr. The implication of this is that it is difficult to keep the discharge from following the vacuum line if the line is at a low pressure<. The solution is to put the discharge regulating valve near the vacuum pump so that the line from the laser is at the tube pressure. Then making the vacuum line smaller and longer than the laser tube will do the trick. Note that SMALLER tubing has MORE voltage drop. Since we're far from the molecular flow region, simple small tubing can be used. This is counter to intuition and conventional wisdom for vacuum systems where larger hoses are almost always desirable.
(From: Steve.)
This is interesting. However, I wonder (publicly) whether it is voltage drop which is the problem. I would have though that breakdown voltage is the key statistic.
As everyone knows, it takes a significant 'overvoltage' to start a gas discharge. Once the gas discharge starts, then the voltage decreases significantly depending on the available current (higher current may cause lower voltage drop; -ve resistance region).
This is why I wonder how you can talk about discharge lengths without specifying the available current.
Without actually trying this, I would have thought that a narrower tube would have a higher current density, and hence lower voltage drop. However, I will defer to your undoubted experience here.
The point I would like to make, is that we are trying to prevent dielectric breakdown in the gas column in the first place (not worrying about how it performs after it's been "lit") - this will be independent of the tube diameter.
It is also good practice to surround the vacuum tube with flexible PVC tubing. This is because some rubber tubes are slightly conductive (e.g. for anti-static properties) and often have a poor dielectric strength. I found this out to my cost when an arc punched a hole in my expensive vacuum line.
Now I use vacuum hose from the auto spare parts shop. The good stuff to get is used in trucks (part of their brake system or something). It's quite stiff, but seems to have better dielectric properties than 'proper' vacuum tube.
(From: John.)
This looks to be a place where cheaper is better. The (relatively speaking) cheap red gum rubber vacuum hose is completely non-conductive.
(From: Steve.)
Measuring pressure:
For measuring pressure, I made a closed end mercury manometer. Normally, filling the tube can be a real pain, but I found that a syringe is good for sucking up mercury. The mercury can be squirted into the manometer tube via some plastic tubing slipped over the syringe needle. I used insulator stripped off wire, however a recent visit to the hospital procured some tubing of the type which is used for epidurals. You can even get the filter section (which I needed, since my mercury had been much loved and was rather dirty). Actually, you can get some pretty neat stuff at a surgical supplier - scalpels, latex gloves, those nifty glass-in-glass syringes to name a few.
I have always had trouble manipulating Pyrex since I only have a bunsen burner. The bunsen can (just) manage to heat tubing so you can bend it. Anything more sophisticated is beyond its power. Thus the traditional technique of melting the end of the manometer closed was not available to me. Instead, I sealed it off with a plastic (acrylic) plug and epoxy. The other end connects to copper capillary. Mercury amalgamates with copper, but I think it will take a long time for the mercury vapour to create a problem. The advantage of the copper capillary is that it can be bent or crimped to create a restriction to air movement. You will appreciate this if your vacuum is suddenly lost - the mercury won't bash out the closed end of the manometer. Since mercury is nasty stuff, enclose the whole manometer in a mercury-proof container.
The first manometer I tried making used silicone diffusion pump oil, however this seems to have quite a strong surface tension against glass (the meniscus is concave as opposed to the convex mercury meniscus) to the point that it would refuse to separate from the end of the manometer. The oil seems to absorb gas and water, which requires that it be de-gassed by heating in vacuo.
(From: John.)
You might want to take a look at the DC-704 silicone fluid. It is much thinner than the DC-705 that many people use. It still sticks to the top of the manometer but a short blast with a heat gun or hand torch solves that problem. I like the expanded scale that the fluid (specific gravity about that of water) gives over mercury.
(From: Steve.)
Cooling:
I didn't want my laser to be tethered to the nearest cold water tap, so I decided to construct a recirculating coolant system. I found an excellent pump in the form of a bilge pump (used by boating enthusiasts). These run off 12 VDC and generate enough pressure head (centrifugally) to raise water from floor to table height. The cost is about $30. The 'tank' is a photography developing tray (11" x 14"), open to the air. The surface area is high enough to dissipate the waste heat.
Since I was paranoid about water leaking over the high voltage stuff, I overengineered the water jacket. It was made of aluminum tubing machined and then shrunk fit into end caps. The end caps included O-ring seals for the Pyrex tube to slip through. The cathode end was directly connected to the water jacket (and this is earthed for safety), so that maximum cooling would be available for the cathode (which tends to be the hot end). The anode was insulated from the water jacket assembly by 3 mm sheet acrylic, caulked in with silicone sealant. The laser assembly was actually supported by the water jacket, which is quite aesthetically pleasing (in my opinion).
Water circulates through the jacket, pumped by the bilge pump. I got fairly good at making hose barbs, since there is quite a lot of tubing by the time the whole thing is completed.
I used acrylic all over the place at the anode end. With the water cooling, there is very little heat.
Before you fly into a panic regarding heat dissipation, just remember that the dissipation per unit surface area of the tube is not all that high. My tube's waste heat is about 1 W per square cm. Just make sure there are no large areas (say over a few cm^2) which are covered by thermal insulators like plastic, otherwise a hot-spot will develop.
Don't put anything between the tube and the water since the thermal conductivity will be impaired. 2 mm thick glass will easily take 30 or 40 kV so don't worry about dielectric breakdown.
Don't worry about a short (say 3cm) section of uncooled tube at either end. Just let the air flow around it.
The purpose of the water jacket is not so much to get rid of vast quantities of heat; rather it is to keep the tube (and hence gas) temp as low as possible. (Sometimes I wonder if dry-ice/acetone would be an ideal coolant...)
Power supply:
You can never have too much power. If you have a choice of transformers, get the larger one! Throttle it back with a Variac. For twice the electrical power input, you will not necessarily double the output. It depends on so many factors. For the same setup, you may actually get more than double the output because the higher current may mean that you can increase the gas pressure, the tube gain, and hence the output coupler efficiency. Then again, you might get less than double the output if your setup is operating optimally at the 30 mA level.
My power supply is a 600 W microwave oven transformer, multiplied up to about 12kV. I found that at very low gas flow, the thing had about the power of a dead match (just enough power to smoulder a small spot on my much abused pine block). Opening the needle valve and simultaneously increasing the variac (to maintain about 70 mA) really boosted the power output (to the point of punching holes in the pine block). Unfortunately, further increases in gas pressure (and flow rate) caused the arc to go out since the power supply couldn't push the required current. A 15 kV/60 mA NST (900 W) should be good enough for pushing the performance envelope.
Measuring power output:
Everyone wants high power. It's very frustrating if you have no means of measuring the power output. I used the principle of measuring the equilibrium temperature of an object which absorbs the output beam. This is compared with the temperature obtained by the same object heated by a known resistance, when not illuminated. Although very slow to take a measurement, it is at least accurate enough to calibrate a faster instrument such as a pyroelectric or thermocouple detector. The 'object' is a finned aluminum block with a 15 degree conical hole. The hole is blackened by overheating olive oil in it. A 20 W wire-wound resistor is clamped to it (for the reference power source).
My new CO2 laser should produce about 500 W if I manage to solve a few problems.
I'm planning to use a folded tube. Currently, 6 tubes of about 600mm discharge length giving a total of 3.6m. These will fit into a 75mm diam cooling jacket. Both OC and TR will be at one end (the grounded end). The other end will be a fixed cube-corner arrangement, which means that only 3 flats will be required for a total of 6 reflections, and adjustment will not be required. Back at the OC end, 4 additional flats will be required for the beam folding. Pairs of these flats are combined into pre-adjusted cube-edge units, to minimize the required number of adjustments.
Although it is a lot of trouble to build a folded assembly, the shorter overall length (1m instead of 4m) will make the unit much more stable and useful. In addition, the increased cross-section and reduced length will allow the gas flow to be increased. Also, it's much easier to make a relatively low voltage supply (11 or 12 kV) to run the reduced length. Imagine the alternative of 40 or 50 kV for a 4 meter tube!
All this depends on the power supply, which is a major part of the whole project. I originally conceived this power supply as a versatile experimental aid. It operates from 240 VAC at 10 A or 415 VAC at 10A three-phase. The output is adjustable between 1 and 31 kHz (in 1 kHz steps), and the duty cycle adjustable between 0/256 and 255/256. The primary winding of the output transformer (a big ferrite core) is fixed at 50 turns, and the secondary is replaceable. I can put a single turn secondary for 12 V at 500 A (e.g. for welding) or 1,000 turns for 12 kV at 0.5 A (for CO2 lasers). The advantage of high frequency AC is that one can use much smaller capacitors and inductors. I am still in the process of designing this transformer, since the current version has too much leakage inductance.
Everything will be microprocessor controlled, so I can program things like pulse timings, feedback control, and so on.
I'm seriously toying with the idea of designing a more modest SMPS (perhaps 1 kW max output) specifically for laser experimenters. This would have the option of control via a standard parallel port. If I get this designed and working (hopefully in about 6 months), I might set up a kit for people interested in building it.
For the USA, would it be OK to design on the basis of 220 VAC being available, or is that inconvenient? It would complicate the design to be able to handle both 115 VAC and 230 VAC operation. Just dealing with 220/260VAC is much easier.
(From: Sam.)
Yes, single-phase 220 VAC (typically closer to 230 VAC) is usually available but would require running a separate outlet in most cases - no big deal. In the USA, 230 VAC is used for electric ranges/ovens, clothes dryers, larger air conditioners, and so forth.
However, it is a simple matter to operate on either 115 VAC or 230 VAC with a constant 300 to 320 VDC on the main filter capacitors - many SMPSs including large ones do this with a single jumper that converts between a bridge and a voltage doubler:
D1
AC o-----+----|>|-------+---------+-----o DC (+)
~| D2 |+ |
+----|<|----+ | +_|_
D3 | | C1 ---
+----|>|----|--+ - |
| D4 | +--o-o--+ +320 VDC to chopper
AC o-----+----|<|----+ - | J1 |
~| | | +_|_
+-----------|----+ C2 ---
| - |
+------------+-----o DC (-)
OK. I'll try to design with the option of filter caps. Without the caps, the voltage doubler trick isn't available so the driver will have to natively work with either voltage (115 or 230 VAC).
I'm expecting the price to be about $400 for the electronic and magnetic components, but this depends on how many takers there are. Typically, one has to order about 75 power semiconductor devices to get a discount.
I might downgrade the power output to 700 W if it makes a large price difference. After all, this is still a lot more power than your typical 15 kV/60 mA NSX (about 400 W).
Regarding DC output, this is an interesting question. There are basically three solutions:
There are other theoretical solutions, but not practical for such high voltages.
(1) suffers from the disadvantage of power loss in the ballast e.g. my laser required a minimum of 40k at 70mA i.e. power loss of 196W. The advantage is steady current which allows best _tube_ efficiency.
(2) Is electrically efficient, but tube output may be lower than it would be if driven by the same RMS DC.
(3) Would have the same performance as (2), but the cost of the rectifier is eliminated, and the power dissipated at the cathode is distributed over the two tube electrodes instead of concentrated on one end. This is an advantage as far as heat dissipation goes, but requires appropriate construction. Another possible advantage of (3) is easier tube starting, since the (HF) AC is going to ionize the gas by capacitive reactance.
The $M question is just how much difference in output is there between (1) and (2)/(3)? I don't know the answer, but I'm willing to bet that there would not be much difference if the operating frequency was over a few kHz. My guess is that the gas will still exhibit laser gain for a few tens of microsecs after the current drops to zero, thus at high enough frequencies the tube will respond more the the RMS current level rather than the instantaneous level. Please tell me if I'm wrong. If someone has the time, they should measure the output of their tube with respect to the 60Hz AC phase i.e. spin a strip of fax paper (wound around a drum spinning on a synchronous motor) and direct the beam onto it. If the fax paper is completely unaffected at one point it means that the power output drops to near zero at the current- crossing point.
Anyway, if I design this SMPS it will be a straight HF AC output -- you will have to make your own arrangements for rectifying and filtering. To minimize 50/60Hz modulation, the rated output will be obtainable for any input voltage over, say, 50 V with the duty cycle varying to keep constant output whenever the input AC is over 50V. This will provide 'steady' output for 80% of the time with 115 VAC input (or 91% at 240 VAC). The power factor won't be real good, and there's a few technical problems to overcome, so specification subject to change!
I'd like to report a successful power-up on my laser. Some photos of the laser can be found via the Lasergrowing 1 Web Page.
I lined the mirrors up very carefully with a little laser pointer and as soon as I turned on the current, it burned a perfect 1/2" round spot on some paper-covered plexiglass I was using for a target. I then graduated to a red brick, and got a nice round red hot spot the same size. When I put in the 7.5"FL lens I bought, things got really exciting. The brick got white hot when placed near the focal length, and the beam melted the brick to a diameter of around 0.2". This is a bit bigger crater than the 30 Watt CW YAG laser I have at work will do. I then switched to the 3" FL lens, and started playing with popsicle sticks. Waving the stick through the beam at the focal length made a really pleasant buzzing noise with a white hot spot on the surface, and would burn a line through half the stick in 100 milliseconds. I could tell the time because I could see the 120Hz pulse rate burned into the stick! If I moved the stick a little slower, it would slice it right in half. The only problem is that there is a strong jet of smoke directed right back at the lens, and it got it a little smokey. It cleaned right up with some lens paper, but I definitely would need some kind of air blast to protect the lens.
The laser would operate over a very wide range of gas pressure and current, so that part of firing up your laser will not be very critical. It ran fine from ~5Torr to the point where the plasma blinked out, with no huge variations in output. I don't have my peltier power meter set up yet so I couldn't measure the actual power out, but it was quite gratifying. The only disappointment was that the electrodes heat up more than I expected. I am using aluminum cylinders 1"dia by 4" long, and they get hot enough to be uncomfortable to touch after about 5 minutes of operation(Yes, the power was off and the power plug disconnected!). If I want to be able to run continuously, I will have to find a way of heat sinking them or forced air cooling them. I am using a very simple mounting system for the optics, just pushing them against an O-ring with the adjustable optics mount. There is more than enough movement allowed by the O-ring to align the cavity. The mount is mechanically referenced to a 2" diameter steel pipe, so small movements of the plasma tube should not affect the mirror alignment(we'll see). I designed the adjusters with ceramic standoff insulators, so the adjustment knobs are at ground(water pipe!!!) potential.
The cavity is about 1 meter, with a 5 meter radius total reflector, and a 80% transmission output coupler. I am using the gas mix recommended in the CORD information, i.e., 1:1.5:9.3 CO2:N:He at about 0.5 liter/min. I just mixed the gas using partial pressures in a small gas bottle. The He and N2 I got at work, ultrapure stuff, but the CO2 I got from the local paintball shop! The active length is 80 cm, with a 15 kV, 60 mA NSX powering the system with AC. I will try to get some pictures to post sometime in the near future.
I am intrigued by U.S. Patent #4,756,000: Discharge Driven Gold Catalyst with Application to a CO2 Laser where the gold is used to recombine decomposed CO2 in the laser. I notice on my 80 cm, 30 mA tube that the last third of the discharge region is turning whit-ish, a sign of significant contamination according to the patent. This happens even at fairly high flow rates (e.g., 4 L/min). I can think of several ways to get the gold on the tube, but before I spend any time on this I would like to know if any practical systems, i.e., commercial, use catalyst systems like this.
I've been burning other interesting substances (all in the name of science, of course!) I have a piece of alumina insulation, and the focussed beam generates a brilliant white spot and leaves a glazed trail on the surface - that's over 2000 °C, according to the books! I can burn nice round holes through mica. I still haven't attempted burning metal, other that a piece of stainless wire that I just got red hot.
I'm still not satisfied with the stability of my mirror mounts. I have to tweak them often to keep the max power up. I'm trying to analyze whether that is an intrinsic quality of using O-rings for the compliant part of the mount, or if it is sloppiness in the adjustable part of the mount.
In the middle of 1999 I built a small CO2 laser. This construction was similar to the Robert Iannini laser capable of 20 to 30 W.
One month ago I built a bigger and revised slow-flow CO2 laser. Here are some of the main construction details:
At this point any help or idea is welcome. As you can understand, there are no filter capacitor in the above PSU. I would appreciate if someone tell me where I can find 1 uF, 25 kV or something similar inexpensive oil filled capacitor for this function. I have in mind some place to buy such capacitor, but the price is prohibitive for me.
+-----------+ +----+ +-------------------+ +----+ +----+ +---+
| Vac Pump |<-| SV |-+-<-| Laser Tubes |-<-| NV |<-| SV |<-|REG|
+-----------+ +----+ | +-------------------+ +----+ +----+ +---+
| | v | |
+----->----+ | | +-|
| +---+ +---------+ | / \
| |VS |--->---| Control |--------->--------+ | |
| +---+ +----+----+ |Gas|
| | |Tnk|
+---------------<--------------+ | |
+---+
Where: REG = Regulator, NV = Needle Valve, SV = Shutoff Valve, and
VS = Vacuum Sensor.
Because I do not have a proper power meter yet, I'm not sure about the power output I'm getting with this system, but I estimate about 93 W, based on the CORD LEOT's equations for hemispherical slow-flow cavity. In the near future I'll buy a power meter from Macken Instruments. This laser is capable of burning holes in ceramics, plastics, and glass but I have not tried sheet metal yet.
I plan to automate the gas support system with electronically controlled solenoid shutoff valves, so when the vacuum falls below of 0.5 Torr the shut-off valves will open and the gas-mix feed the system with desire pressure. When pump stops, the valves will close, protecting the cavity from overpressure.
Details on my CO2 laser can be found at: Emission Technologies, Rob's company founded to sell CO2 laser plans, kits, and even a complete CNC laser cutting system.
What I wanted to accomplish was to build a CO2 that would be light weight, extremely stable, and powerful enough to do something with. My intent is to build a computer cutting table from scratch including the software which I will write in Borland Delphi 4. This tube puts out over 30 watts (and I have also built a similar 60 watt version.
The current design of the laser head itself is very stable and I am pleased with its performance. You can run it for 0.1 second bursts for drilling, or leave it on for twelve hours at a time. I figured hey, the commercial heads are mostly all glass looking like an overgrown HeNe, so why can't a home built tube use "glass only"as the structural member of the laser. I just used low expansion Pyrex with a relatively thick wall.
As for the machinist part, I have been tinkering for 20 years machining things here and there so I have basically taught myself. This tube was built in about three evenings of sitting at the lathe. Then a couple more nights spent re-machining parts to get the bugs worked out with an isolation issue. The mirror cells are stone cold electrically and this tube is safe to touch, adjust, or hold while running. This was another design parameter.
The cost depends on your stock of raw materials for machining, and your resourcefulness with the other components. The mirrors are the biggest expense for the tube at $50 for the HR, and $150 for the OC. The large Pyrex tubing comes in 48" lengths for about $40. The bore tube is neon sign tubing and can be had locally for $5 per 48" section. Teflon stock is pretty expensive though at $125 for a 2" by 12" piece of bar stock. I probably have $325 invested in the tube. The vacuum pump is a SPEEDIVAC manufactured by Edwards High Vacuum Ltd. in England. They cost about $2,000 new but I lucked out and got this one surplus in new condition for $300. The manometer is hand made from a 1 mm ID glass capillary tube and reservoir of mercury procured from several mercury switches. The gas flow gauge and needle valve are surplus and were cheap. Gas bottle - $125 new from welding supply plus about $30 to fill it with the laser mix. Gas bottle regulator - any inert gas regulator adjustable from 0 to 20 psi will do and they cost about $120 new.
In doing the math and assuming the use of a 40 cubic foot Bottle of gas mix (8" diameter x 18" tall), you should get about 18.9 Hours usage at 1 L/minute flow rate. The cost to re-fill the bottle is around $25.00.
The coolant system consists of 1 gallon tank with $18 bait well pump (700 gph) from Wal-Mart boat section running from a 12 VDC supply, and a copper heater core from a GM car with 7" diameter muffin fan. Fan was $25 but I don't know on the heater core as I had it already. I suspect they are $80 - $100 new. Power supply is a 15 kV @ 60 mA neon transformer - $150 new, with 6 amp Variac $25 surplus, homemade bridge rectifier from 60 1N2007 diodes - about $25 worth, and eight 2000 VDC @ 1.0 uF caps in series for filtration of the output - $32 surplus but could be $400 if bought new (High Voltage Caps are Expensive). Ballast resistor is (now) ceramic 100K @ 250 Watt - $18 new. A 0 to 60 milliamp gauge at $10 completes the power supply.
As you can see I have about $1,200 in this one but the costs could be a lot more if you are not a good scrounger of surplus parts.
I will keep you informed on the results of the folded resonator design. Once these are all de-bugged and I get my AutoCAD Drawings updated, I may offer partial or full kits of the laser head, and possibly a complete roll around enclosure that holds all of the support equipment in a nice neat package.
(Update as of Summer, 2001)
I am now offering a kit of parts for the 35 W and 60 W laser tube assemblys as well as completely assembled units. There is more info at my Web site, above.
My first hands on experience of flowing gas lasers was gained many years ago, in the biomedical field, a 35 watt flowing gas surgical laser. This used pink laser mix 1 and had lots of solenoids and valves. It was entirely microprocessor controlled.
We were all amazed by the quantum mechanics of the whole process, the manual was excellent being almost an instant course on the history, development, problems and applications of cutting by this invisible light. Also great emphasis was placed on the various safety hazards that exist from such a device.
At first glance, either a relatively cheap glass sealed CO2 laser or perhaps the latest small industry OEM units could be integrated along with our computer controlled XY tables with excellent results. Although the cost of these range from moderate wallet pain to 2nd mortgage level, it would be difficult to justify such a sum for casual use.
Although someone who needs a laser would choose the above path, my interests mainly lie in the educational and fun value of building a prototype tube. This actually works, a laser that is user serviceable would be handy for some, but this would not replace conventional tooling for most folks needs.
This project immediately ran into difficulty during the on paper design stage. It became clear that the resonator HR and OC would either be a very expensive custom built design or perhaps a cheaper stock optic.
Although I originally wanted a ZnSe OC for ease of HeNe laser alignment and transmission of IR, partial coating to a specified reflection was prohibitive. Ge was easily secured so the basic natural 50% reflective OC was one concrete equation.
The HR is currently a flat 1st surface enhanced aluminium mirror. I am keen to see if a copper mirror, gold plated, would work as we could experiment with different RoC values in-house. A true dielectric 10.6 micron HR mirror could then be possibly be custom made with some confidence?
The requirement of a reasonably high gain resonator kept the discharge tube diameter down to 10 mm, 600 mm in length. This proved easy to start with a 10 kV, 100 mA neon sign transformer and Variac. I gather that this is the maximum allowable U.K. rating for double insulated neon signs.
I was keen to keep the plasma discharge within the cooled area of the resonator and that the resonator should be easily removed / replaced along with the turned Al electrodes and sub assembly.
As I already had CO2 and N2 in-house, it was easier for me to rent a high grade He cylinder and use a rotometer mixing manifold for my laser mix. This seems best at B.O.C. laser mix One/laser mix Two ratios. Premixed gas would remove any doubts here.
A two stage rotary pump seems to be easily capable of providing a good vacuum level. Fresh laser mix gas flow rates of around 2 to 4 litters per minute seem OK. The vacuum pump is therefore required to cope with this virtual leak and keep within the current 20 mbar vacuum envelope.
Cooling was provided by a closed circuit water circulation.
Again options exist for either using a copper turning and focusing mirror or perhaps more a more desirable specialist 45 degree mirror and a converging lenses with gas assist. Again in house milling and turning would probably do the trick for the mounts at least.
I am very pleased that the laser actually works, as a garden shed design. A short video clip is available on the Compucut forum (take link below), very early days yet. Here is a photo of Iwan's Home-Built CO2 Laser in action.
But the infrared output has got me excited enough to continue this labour of love.
To ensure that results are repeatable this basic design will be replicated twice. Luckily I have enjoyed good relations with several U.K. based specialists, neon sign power supply vendors, optical technicians and Master glass blowers. It was an important factor that parts used would be available at the same specification for further developments.
I really appreciate the kind help that I have received, especially the companies that took me under their wing, donating data, sending one off free end of line samples and in some cases steep discounts on current production line equipment.
I am especially indebted to Sam's Laser FAQ and David Knapp's notes.
As you can see the design has evolved, the glass rod spacers did not work too well on the 1st attempt.
Hopefully this type of laser along with my friend's computer controlled CNC XY table setup will be a great boost for my garden shed!
I have my own CNC mill and lathe from Compucut here in Coventry.
(From: Steve Hardy (hardy@keeper.stortek.com).)
The most-often-asked questions about CO2 laser optics are the required reflectance of the Output Coupler (OC) mirror and the radius of curvature of the High Reflector (HR) or resonator mirror. The following sections should help you answer those questions.
Refer to the simplified drawing below while reading this dissertation. :)
HR <----------- Bore Length (Lb) ------------> OC
(Curved) ___________________________________________ (Flat)
|( ____ _________________________________ ____ || ===> Output Beam
| | (R reflectivity)
<---- Discharge Length (Ld) ---->
<------------- Resonator Length (Lr) ------------->
The optimum value for the OC reflectance is given by:
Ld
R = 1 - ---------
500 * D
Where:
For example, my tube is 1,200 mm long with an inside diameter of 15 mm. The desired reflectance is thus 0.84 (i.e., 84%).
In practise, there is a bit of latitude, since this formula is for a 'typical' axial flow. My OC happened to be 75% reflective and works fine. Given the choice, go for slightly lower reflectivity since it allows for 'future enhancement' by the radiation-crazed power-addict.
Note that for longer and/or thinner tubes, the reflectance needs to be lowered since there will be higher gain. It is not obvious that thinner bores lead to higher gain, but this is due to the fact that thermal conductivity from plasma to tube wall is enhanced with the smaller distances. It is most important to keep the gas at a low temperature.
Here are some examples (for relatively long tubes but you can recrank the math for your tube length):
Assuming the 15 mm ID, the formula above results in R = 75%, 80%, and 85%, for tube lengths of 1,875 mm, 1,500 mm, and 1,125 mm, respectively.
However, I would probably recommend a slightly lower R like 75% where 85% falls out of the calculations because:
Single mode (TEM00) operation is probably the most desirable, even if not always as powerf