TJIIRRS: Number 15

(23 July, 2007)

Although various people have made excellent holograms with ordinary laser pointers (which turn out to have surprisingly good coherence length; see below), there is an issue; because the process is extremely sensitive to vibration (this is, apparently, especially true for reflection holograms), you want to keep your exposures brief. In order to do this, you need to increase the amount of light that goes through your setup. If you have only a few milliwatts to start with, there is a limit to how far you can get by minimizing losses.

[[When I lived on the plateau in what was then Issaquah, Washington (I believe the plateau part of the area is now a separate town called Sammammish), I had The World’s Most Stable Basement Floor. I could make beautiful holograms with a 1-mW HeNe laser just by tossing a piece of plywood onto the carpet and putting the optics in place. The only thing I had to worry about was the furnace; if it was running, I would wait until about 60 seconds after it turned off, to give the air in the room time to settle, before beginning my exposure.

To give you another index, one night we were down “in town” at the supermarket, when there was a 4.8 or 4.9 earthquake. This is not particularly horrendous; as I recall, one can fell off a shelf. Still, it is not something you are going to overlook — the whole building was rocking and swaying, and at least one person screamed. They told everyone to leave the store, no surprise; we went up to the supermarket at Pine Lake, about half a mile from the house, and we had to tell people that there had been an earthquake, because nobody had particularly noticed it!]]

That was then; this is now. My current setup is literally 35 feet away from some rather active train tracks. Even at 2 AM there’s a train about every 10 minutes. To be sure, there are many gaps of half an hour or longer, even during the day; but there are also times when it seems like there’s a train every minute or two, and there are definitely times when two trains pass each other here. (I haven’t tried to work out a schedule; perhaps I should.) Also, we are on sand and clay here; I can sometimes feel it when a truck goes by on the street in front of the building, and that is a lot farther away than the tracks are.

In order to minimize exposure times, I decided to see how well I could do with a diode laser of somewhat higher power than a pointer, which may not have characteristics that are quite as nice; in particular, I worry about the mode structure of the beam and also the coherence length. (This is somewhat lengthy, and not necessarily of crucial import, so you can skip ahead if you like.)

My first move was to acquire a device. I purchased a 100-mW 645-nm laser diode module (on eBay, from Scientificos, an excellent vendor from whom I have purchased other items in the past). This device arrived with appropriate cautionary notes suggesting that even a reflection of the beam could damage your eyes, and I am being careful with it.

With the laser in hand, I tried putting an external mirror on it, in an effort to create a second cavity, into which I figured I would put an etalon to try to narrow the bandwidth. About the only things I had that I thought might be viable were old HeNe mirrors. I didn’t want a flat, but I have several of these mirrors, and I found an output coupler.

It is nearly impossible to match the beam divergence with a mirror of any reasonable curvature, particularly a short-focus mirror like an OC from a small HeNe tube, so I decided to turn the mirror around and effectively make an unstable resonator of a rude sort. This vaguely worked, but the power on the far side was much lower than the power coming out of the diode, and that was clearly not going to work very well, so I decided to try an etalon by itself, and not bother with the mirror.

The results were considerably better.

Here is the laser module temporarily lashed to a rail, with a small etalon in front of it:

When I get the etalon aligned nicely with the laser, I notice that at low and moderate power levels, I actually see a brighter spot with the etalon in place than I do without it. Here are two comparison pairs, one at low power and one at moderate power.

It’s a little bit hard to see in the photos, but the difference is readily apparent to the eye, and my power meter gives me readings as much as 10X higher. (The meter is on its 633 nm setting, but it is old and out of calibration now, so I am taking only relative readings from it.)

At relatively high power, OTOH, the power clearly is not increased (and sometimes appears to be decreased) with the etalon in place. I’m not sure what gives here, but I will try to find out.

There are also differences in the appearance of the speckle, but they are very difficult to photograph and not necessarily easy to interpret. Here are two speckle photos, taken with the camera (Sony DSC-R1) at full telephoto, and manually focused as close in as possible in normal mode, to make the spot larger. (Note that laser speckle is an interference pattern that is present in space; what you see when you observe speckle is that pattern as it exists on your retina. What you see here is the pattern as it existed on the sensor of the camera. It doesn’t really even make sense to talk about focusing on speckle, as it is not an image in the ordinary sense of that word, so there was no particular reason to focus the camera on the spot on the paper target.)

I suspect that as my next step I am going to have to build some sort of apparatus that will allow me to measure the coherence length of the laser, without and with the etalon in place. It would be nice if the module has sufficient coherence length as is; I would be looking at an exposure of less than 1 second for a 4"x5" plate. If the coherence of the module is not good enough on its own, I just hope I can get a decent enough beam out of the combination of the module and the etalon. Either way, I need to check.

A Michelson interferometer may be the easiest way to do this, but I will be checking Sam’s Laser FAQ, a good source of information about many aspects of DIY lasers, first. (Although there are some places on the FAQ where people make statements that are clearly wrong, Sam’s own info and plans are generally excellent, and quite a few of the other folks are fine as well.)

Sure enough, there is some discussion about the coherence length (and time) of semiconductor lasers, and they refer to US Patent #4,907,237, which is about stabilizing the things.

(evening of 19 August, 2007)

I have started to set up a Michelson on the bench. It is still somewhat loose and informal, but may serve well enough. If I find that it is too loose, I will tighten things up.

The beamsplitter I am hoping to use is rated about 70/30, and is apparently plated with chromium and gold; when I get a bit further along I will measure both beams to see whether it works well enough. If not, I will find a different one.

I have a pinhole and a microscope objective, with which I hope to make a spatial filter that will also expand the beam; but as of this writing I am not entirely certain how to mount either of them. (There are special mounts that hold one of each, specifically for this purpose. I do not have such a mount, and I don’t think I have the tooling to build one.) My reason for expanding the beam is that if I just aim the laser into a Michelson, part of the beam will be reflected back into it by the mirrors, and will almost certainly change the performance. If there is a 200-μ pinhole in the way, very little of the beam is likely to get back into the laser. The side advantage is that the spatial filter will also give me a cleaner beam.

(Some time later)

I have found a 1" mount, which will take the microscope objective if I’m careful. It had only one retaining ring, but I found a ring that is just the right size to glue onto one face of the plate as a retainer. (That is, the diameter of the open area on the plate is 1"; the inside diameter of this ring is perhaps 22 or 23 mm, so it only obscures a wee bit at the edge, while still providing a tiny shelf.)

My pinhole (see sidebar, below) is not held firmly in place, so my results are a little bit wobbly; but I now have fringes on a paper target, and the mirror paths are at least 10 cm different, so if I can stabilize everything I should have no trouble making some sort of hologram. Next step, I think, is to order some plates.

Here are some photos. First, the Michelson setup. I have attached the etalon to the aluminum plate, but because I want to be able to make measurements both with and without it, and because the black baseplate comes from another device (and I don’t want to make holes in it), I have not attached it firmly. The tungsten [or tungsten alloy] weight holds it nicely in place.

Second, a detail showing the spatial filter. (No rude remarks, please. I just wanted to see whether I could make any fringes happen, and I don’t have a mount for the pinhole, so I just put it in place. It’s the crumpled piece of aluminum foil, in which I have actually made a dozen pinholes.)

Third, the setup in the dark, with the laser running.

(20 August, 2007)

I ordered a box of 2.5" square BB 640 plates from VinTeq Ltd last night. These are manufactured by Colour Holographic, in England, and are clearly very good. I hope, eventually, to make my own, generally following this article, but that will have to wait until I have time to mess with it. (The first author of the article, btw, Jeff Blyth, is one of the developers of BB plates.)

This morning I moved the mirrors around. I am still seeing fringes with more than half a meter difference in pathlength, both with and without the etalon, though it seems to be harder to get them with the etalon in place. It begins to look like the main use of the etalon may be simply to get decent power output at slightly reduced electrical input. (But see below.)

(An hour or two later)

After some more fussing, I have discovered that the etalon can, indeed, stabilize the laser, but that it only seems to work at some power levels and some alignments. With the etalon in place and everything set correctly, I can see that sometimes the laser locks in (the fringes become very distinct) and sometimes it loses lock (the fringes lose contrast, and sometimes disappear entirely). Even when it is locked, it clearly does a lot of “mode hopping”. The changes are rapid, on the order of seconds, which means that the laser is not yet stable enough for holography. Here are photos, taken within a minute or two of each other:

I will have to work some more on this, and I may try adding an output coupler, outboard of the etalon; my current idea is to try something with very long radius of curvature, and roughly 50% reflectivity. I also think I will have to operate the laser from batteries instead of a power supply that plugs into the wall. Just hope I can get a viable power level with batteries. (Not much choice of voltage.)

If you cannot get an acupuncture needle, try sharpening a sewing needle (see my page on building a Voss electrostatic generator for a sharpening technique) and using that. Just don’t press it down too far or too hard. You will probably want the bottommost pinhole, as that will be the smallest.

Here is a short AVI movie, showing the fringes being fairly steady for almost 30 seconds. I have changed the configuration several times since then, trying various orderings and combinations of layers, in an effort to get some understanding of what works and what doesn’t. My current stack (which you should not necessarily regard as final) is composed of a set of chipboard shelves that are 2'x4' (a little over 60x120 cm). Here’s a listing of what’s in it, with “================” to represent the shelves:

1. A 2' x 4' piece of thin steel sheet, which serves as the top surface. (This is new as of the second week in October; it replaces the 2' x 3' piece of phenolic circuitboard I was using earlier. See the text for more about this issue.)
   
   
2. ===================== (The top shelf) =====================
   
   
3. A thick cotton bathmat (I actually had to use two, as the first one was about 2' x 3'. The second one isn’t quite thick enough, but is perhaps a bit too thick when I fold it over, so I may eventually look for one that is a better fit.)
   
   
4. =============================================================
   
   
5. Another folded quilt, this one much thinner and more firm than the one at the bottom.
   
   
6. =============================================================
   
   
7. Six pieces of actual Sorbothane, Shore 30 hardness (that’s the soft and squishy version of Sorbothane; the harder version is Shore 50). These are 4"x4"x¼" (~10x10x0.6 cm). Each of them sits on a slightly oversized piece of hard white packing foam, also about ¼" thick.
   
   
8. =============================================================
   
   
9. A folded piece of canvas, probably 8 layers thick. (This is probably a piece of bedding, but I put it into the stack without completely unfolding it, so I’m not entirely certain. It definitely came from the bedding area at one of the local thrifts. I wanted canvas because I noticed that a canvas duvet cover I have stops sound a lot better than a fluffy quilt; it is important to have several different kinds of layers, to stop different frequencies of vibration.)
   
   
10. =============================================================
    
    
11. A triple layer of openweave nonskid (rubberized) drawer liner. (I think of this as “poor-person’s Sorbothane”. It is probably fairly good at damping vibrations.)
    
    
12. =============================================================
    
    
13. 5 very small bicycle innertubes in an "X" pattern. They are about 12" across (~30.5 cm) and just over 2" thick (~5.4 cm). I have not inflated them fully, as they would be too bouncy that way. In fact, they are too bouncy even when partially filled, so I had a bright idea and covered them with a folded blanket, which seems to quiet them down quite a bit.
    
    
14. =============================================================
    
    
15. A folded quilt or comforter of a rather ordinary sort, perhaps an inch thick when unfolded. I think this is folded in thirds and then in half; but it compresses under the weight of the table, so it is probably no more than 2" thick now, and possibly even somewhat less.
    
    
16. ======================= (The base ) ===========================

(Early October, 2007)

I am coming to the conclusion that chipboard shelves may be okay as interior layers, but I may possibly need something stiffer on top of the Sorbothane, and I almost certainly need something stiffer as the working surface on the top of the stack. (I tried a piece of phenolic laminate that I have here; it is about 2' x 3' and 5 or 6 mm thick. That helped, but a few days back I found a thin steel sheet that is just over 2' wide and very nearly 4' long, so I swapped out the phenolic. I think the steel sheet is almost as stiff, and it gives me more working area, though there is a rectangular hole in it near the laser end of the table, which may eventually prove to be annoying.)

In the meanwhile, I have not been entirely satisfied with the beamsplitters I’ve been using in the interferometers, and I have ordered a cube beamsplitter from The Surplus Shed. I have also started lightproofing the room, so stray light won’t fog the plates, and I have bought some ancient shutters on eBay, to see whether I can use one to control the exposure. (Two of these arrived today, 28 September, and they both look viable.)

I am getting my holographic plates from Dr. Joachim Vinson at VinTeq, and my photographic chemicals from The Photographer’s Formulary, an excellent source.

(01 October, 2007)

Also in the meanwhile, I decided to separate the laser from the table in terms of stability, to improve my understanding of what’s going on. I took a single-longitudinal-mode stabilized HeNe, put it on the table, set up the interferometer (just a Michelson this time), put the output through a concave lens, and directed the result to a piece of white paper. Once the laser locks, which takes about 20 minutes, any motion of the fringes is essentially guaranteed to be a table issue.

It took about a day and a half for the table to settle down (I had ripped it apart to put the white foam pads under the Sorbothane pads), but as of today the motion of the fringes is so slow that I won’t show you a movie — it would be too boring. I did learn some crucial things, though. The first is that it does take the table that long to settle after I do any major work on it. The second is that almost any air motion is enough to disturb the fringes, so I am going to have to build a box that completely surrounds the table. I am currently expecting to attach the shutters to the rear wall of this box, and to make at least the front and possibly the top out of cardboard so I can move them aside easily when I need to change setups. The back and the two ends, however, will definitely be wood, probably plywood because we have lots of it here.

I say shutters, plural, because I want to have a light-block to prevent stray light from reaching the plate, in addition to the shutter that controls the exposure. The light-block can be very simple, but it also has to be very effective. There isn’t much room for this, however, unless I do something to collimate the beam; the divergence is huge, probably about 30°.

(15 October, 2007)

I have now had a lot more experience setting up both the table (which I have modified several times) and the interferometer (which I have had to take down every time I rebuild the table, and which I also tweak at other times). I have become firmly convinced that even fairly heavy mirror mounts are not as stable as I might like to believe, and this afternoon, with some help from our Range Safety Officer, I used Museum Wax (which is probably very similar to Quake Wax) to stabilize the pieces of the interferometer. About the only part that I didn’t wax into place was the beamsplitter cube (which, btw, works pretty well). The results were fairly dramatic: I can barely see a change in the fringes when I stomp on the floor a few inches away from the table.

On the other hand, my feelings about air currents are, if anything, stronger than ever. If I even breathe toward the table, I can see the fringes move. Some of this appears to be thermal; but in plain point of fact it doesn’t really make much difference; the table simply has to be protected against air currents. Period. I have almost all of the pieces of a box to go around it, and I will be putting that together soon. I still need a lid, and I need to hang a hook from the ceiling so I can work on the setup without having to hold the lid up by hand.

(16 October, 2007)

Last night I latched the enclosure pieces together, and there is now a box around the table. I also tweaked the table layers today (and changed the description above so that it is current). I am currently using the phenolic board and a largish cardboard box as a lid, so I can easily get inside.

The table still dances madly when a train goes by, but it is otherwise relatively calm and quiet, despite the fact that it has had only about 30 or 45 minutes (as I write this) to settle down. My next issue is the laser; it seems to mode-hop for quite a while after I turn it on, and I may eventually decide that I need to stabilize its temperature actively. Fortunately, I have a spare Peltier cooler here. Unfortunately, active stabilization is somewhat painful and tweaky; it would probably take me some days of frustration before I got to anything viable.

(About an hour and a half later)

In an effort to avoid that path, I have just reoriented the diode, rotating it 90° to give it a lot more area in contact with the heatsink. I had it pressed into the corner of the heatsink, like this, with the beam coming out of the screen toward you —

        |
        |  _____________ (pressure plate)
        |/    \
        |      |  (diode)
        |\____/_________ (heatsink)

— and that may have provided insufficient heat conductivity. (Notice the fact that there are only three small points of contact, and one of those is on the pressure plate rather than the heatsink. My apologies for the ASCII-pic; if I have time, I may try to draw a better diagram of this.)

It is now sitting in a hole in the heatsink. The base is flush against the surface, with a bit of heatsink grease on it; and the pressure plate (also with a bit of grease) is pressing it against the heatsink. The beam emerges through the hole in the heatsink (a red arrow points toward the laser in the first picture) —

At this point it seems to settle down reasonably well after 20 or 30 minutes of operation, and once I figure out a way to introduce a shutter I may actually be able to make some holograms. There is some chance that the laser will end up outside the box, with the shutter against the wall, as that would be relatively easy to implement. As John F. Pecora has clearly demonstrated (Click the link labelled “Proving Fringe Stability with Laser off Table”), the requirement for stability begins at the point where the input beam is divided into object and reference beams. (In the case of a single-beam hologram, this actually occurs at the emulsion; for commonplace 2-beam setups, of course, it is at the initial beamsplitter.)

(Note, added on October 19, 2007: I should point out that if you want to be able to detect mode-hopping and other instabilities in the diode, you have to have different path-lengths in the two arms of the interferometer. Remember that if the arms are equalized well enough, you can even get fringes with white light.)

In addition to working out the details of the shutter (and isolating it from the table), I now have to generate the other optics and bits I need, for example a plateholder. Here is an early stage of my initial thought about a holder for Denisyuk-style single-beam work with 2.5" square plates:

The body of the holder is a sample piece of composite that is intended for people to build porches and decks from. This stuff, in larger pieces, would probably be used for [vertical] posts. I have added a rectangle of plastic pieces, to create a central hole that is a little less than 2.5" square, and I’ve added two corner pieces to keep the plate in place. I will probably chop a hole in one side and put a mirror at 45° inside, so I can keep the laser separate. Hanging lead-acid batteries on the sides, to add mass and stability, seems like a good idea. I also intend to paint this entire object matte black, inside and out.

(19 October, 2007)

The holder is now finished, at least to first order. The hole in the side that admits the light is fairly rough, and I will probably have to enlarge it at some point; but it will do for initial testing.

(12 December, 2007)

In examining Denisyuk-type setups made by other folks, and thinking about it, I realized that this holder is flawed; it would require the viewing beam to go straight into the plate, from the position of the person doing the viewing. That’s not really hard (this is what beamsplitters are for, right?), but it means that there would be a nasty bright reflection off the front of the plate. I will be reworking the holder, so that it brings the beam into the plate at a reasonable angle.

(12 December, 2007)

I have been asked to consider doing a holography demonstration. Because the venue is down in DC, it seemed like a good idea to have a table that is reasonably portable, so I built one. It is 16" square, and uses a single 14" bicycle innertube. It also uses two slate tiles, which are more massive than the chipboard shelves I used for the other stiff layers.

I am currently looking into a nifty shock-absorbing material called d3o. Here is a YouTube video about the stuff. I have no idea whether it will absorb the small vibrations that are my concern here, but it appears to be very good at stopping large shocks.

Needless to say, btw, there is another way to get short exposures: use a pulsed laser. If you can get sufficient energy in a pulse that is no more than about 1 μsec long, you can even do holographic portraiture. The next page [or set of pages] in this series is about my attempt to do that with a flashlamp-pumped organic dye laser, and possibly also with a small Nd:YAG or Nd:Glass laser and a doubling crystal.

Back to the Index

Home

Contact Information:

My email address is a@b.com, where a is my first name (just jon, only 3 letters, no “h”), and b is joss.

My phone number is +1 240 604 4495.

Last modified: Wed Dec 12 02:04:42 EST 2007