TJIIRRS: Number 5C [New] of an Ongoing Series;
“Theorie und Praxis IIA”:
Revamping the “DKDIY” Laser
(15 August, 2006, ff)
This page details the construction of a nitrogen laser that is a follow-on to the “DKDIY”design I published here a few months ago, along with a “How-To” page. Because this material is being written substantially as a historical track of the project as it is taking place, it is not necessarily organized logically. When the design is fully stabilized I will try to provide a “How-To” page for those who want to build a laser of this type.
(Note, 2006 September 27: Between the “DKDIY” laser and this “DK-Plus” laser, I experimented with a larger design, which operated, but not at the performance level I had expected. This appears to have been caused by several factors, some of which I may explore [and, I hope, correct] by returning to that laser and rebuilding it, now that I have this one working well.)
(Note, 05 October, 2009: I am reworking this laser, and I hope to get somewhat better performance from it than I originally did. There are a number of issues involved in the rework, which I discuss at one or two places on this page, and in a second page specifically devoted to it.)
I would strongly suggest that you read through at least the early sections of this page and all of the following page before you attempt to build one of these, as that will help you avoid some “gotchas” that I encountered.
!! CAUTION !!
This laser uses high voltages, and capacitors that can store lethal amounts of energy. It puts out an invisible ultraviolet beam that can damage your eyes and skin. It is important to take adequate safety precautions and use appropriate safety equipment with any laser; but it is crucially important with lasers that involve high voltages and/or produce invisible beams!
Origins and Design Rationale
The first version of this laser was intended to provide performance at least as good as that of the classic Scientific American “Amateur Scientist” nitrogen laser design, but using “doorknob” capacitors instead of circuit board, and with a triggered spark gap as a switch, rather than a free-running gap. It is a “Voltage Doubler” circuit, often mistakenly called a Blumlein. I chose the doubler circuit because it is relatively easy to construct, and because it parallels the circuit of the SciAm design.
That first “DKDIY” laser used a dozen laser-grade SrTiO3 doorknob capacitors. It reached threshold at about 12.4 kV, and developed a little over 100 kW at about 20 kV. The channel was 22 mm across and about 45 cm long, and the electrodes were pieces of extruded aluminum carpet edging.
This revised version will use 16 doorknobs, and will have electrodes a bit more than 80 cm long, with capacitors along the middle 45 cm or so. I originally intended to use a piece of brass shim stock as the cathode; shim stock has a fairly sharp edge and so should provide some preionization by generating corona early in the discharge cycle. It’s a bit thin in comparison with the anode, but may work. If not, I can always try a packed-blade cathode, or add some extra preionization.
(Note, added later: I soon abandoned the idea of a thin sharp-edged cathode, and shifted to two identical electrodes.)
My original intention for this enlarged version was to use carpet edging again, at least for the anode; but when I examined the pieces I had on hand I discovered that they were not very straight. Straightness being a real issue, I changed my mind and turned to something that does have a straight edge: an aluminum ruler. The ones I’m using were originally 4 feet long, but I have cut them down to 32 inches, a reasonable length for this head. They are 2″ wide, which fits well, and 1/8″ (~3.25 mm) thick, also very reasonable, though that is significantly thicker than the previous electrodes, which had a cylindrical edge only 1.5 or 2 mm in diameter. My hope is that the capacitors I’ve added will be enough to let me pump the additional channel volume. (It is important to deposit sufficient energy into the discharge, in order to be sure that the laser will operate well above threshold. Typical high-performance nitrogen lasers seem to dissipate at least 40 joules per liter of active discharge, and some exceed 100 j/l.)
Following an idea developed by Jarrod S. Kinsey, I am constructing the sidewalls and spacers of this new head from wood, a construction material that is widely available, extremely tractable, and relatively inexpensive. (I may provide windows through the sidewalls to permit observation of the discharge in the channel.) Wood tends to be slightly conductive at high voltages, but I will be using walls that are varnished, which should reduce the conductivity somewhat. If some conductivity remains, there is a small chance that it may provide a bit of preionization, in much the same way that a semiconductor plate does.
(Addendum, 27 September, 2009)
Most wood is a bit porous, which is a problem. (It turned out to be a problem even with the varnished yardsticks I used in the head of this laser.) Consider a nitrogen laser running at 50 Torr; if there is a leak that admits 2 Torr of air, the gas mix contains more than 3/4 of 1% oxygen, which is too much for optimal performance. (A small amount of oxygen is okay, and in fact at roughly 1/3 of a percent you may see slight improvement in performance. Beyond about half a percent, however, oxygen degrades the performance of the laser.)
If you use wood, make sure that you either coat the interior surfaces with something that is relatively impervious to air, or soak the wood in something that will solidify and make it impervious. One way to do this is to mix epoxy with isopropyl alcohol (at least 91% pure; 99% pure is better), and brush it on. You probably want the epoxy to be just liquid enough that it brushes reasonably easily; if it is too dilute it has a tendency either to fail to cure, or to cure to a soft and rubbery condition. After the first coat of epoxy has had time to cure, repeat the process until it no longer soaks in; this will almost certainly take several coats. When all of the epoxy has fully cured, the wood should be sealed well enough to be usable.
There are certainly other ways to accomplish this, but you will want to think it through carefully. Some paints, for example, continue to emit solvent vapors for months. These may (or may not) interfere with lasing. There is only one real way to find out; and if the answer is ‘yes’ you will probably be obliged to rebuild the head, which is (believe me) a real annoyance.
Circuit and Physical Layout
Here is the circuit diagram of the laser:
The capacitor labelled “Start Cap” is present mostly in order to produce a current of a few dozen Amperes in the spark gap very quickly when it is triggered, to help develop a substantial conduction channel in it. The manufacturer recommends pushing at least 10 Amperes through the gap to get it to switch properly. This probably takes only a few dozen pf, which can be furnished by a rather small doorknob; it is, though, important to keep the connections fairly short and the inductance down, as you have only a brief time in which to accomplish the job: a good spark gap should switch in a few tens of nsec. (Please note that although it is possible to make subnanosecond spark gaps, the designs I’ve seen were pressurized to about 1500 psi and were built into cylindrical transmission lines.)
I have shown a charging inductor across the channel, but a charging resistor works better in some designs, and I will probably try both to see which is appropriate for this laser. I have also shown a small capacitor that connects to a dot near the cathode. The dot represents a thin wire that is strung along the entire length of the head, and serves to preionize the laser. This is discussed in the text, and is the initial configuration I’ve chosen to try; I may move to a different preionization method later.
(Note, added much later: I did. See below.) Last modified: Sun Dec 18 00:06:08 EST 2011