Turning A Magic Wand

January 31, 2013 by Kurt

In the run up to Christmas our family went to Dickens Faire, and my son really wanted to buy a wand at the wand shop. I told him we could make one at home. Now that the Christmas rush has passed we’ve finally gotten around to making one.

We chose a length of dowel rod, and I marked the hand-grip section out with a pencil. Then I put it in the lathe and turned it out freestyle. We did it out in the driveway to keep my shop from getting full of wood shavings. I really should put some roller wheels on the end of that lathe. It’s kind of small but still not much fun to haul out from under the workbench and lug out to the driveway.

My son tries the wand’s grip to test the comfort level.

I couldn’t find all my turning tools, but was able to track down a big gouge and I made do with some chisels I had around. My son seemed to enjoy the process. He was amazed at how the dowel seemed to stand still after the not-so-centered wobbly parts were turned down and it was running true. The wand is not going to win any awards but he can hardly put it down.

This summer we dug for Herkimer Diamonds and we’re planning on gluing one to the tip of the wand. I also will try to make a secret compartment in the handle by cutting that trailing ball off, center drilling, and re-attaching with a dowel. Lord only knows how I’ll hold the wand in the lathe for center drilling. I guess I should have though of that first.

Pastry Cutter Final Results

January 26, 2013 by Kurt

After finally getting the last parts cut on Dec 23 I had to go into a fabrication frenzy. I was constantly ducking out into the garage to do the next round of glue and/or paint. One of the real tricks with these cutters was getting the push plate positioned really accurately inside the cutter walls. There was no positive alignment when I glued the cutter walls to the top plate. If the push plate was off by much the plate would bind against the wall and stick. Here you can see the system I came up with. I glued up everything else, and positioned the backing plate on a scrap rod. Then I just put in a few drips of solvent on the backing plate, and plop on the push plate. I had enough time to position it exactly and then let the solvent do it’s work. I left the protective paper on the under side of the backing plate so there was no danger of a stray drip of solvent fusing the backing plate. This system worked great, but it did mean that the parts weren’t interchangeable. That really didn’t matter much.

I cut some springs/push rods. I needed them to be just the right lengths so the spring would still exert some upward pressure when fully up, but would bottom out before the push plate could be pushed all the way out of the cutter housing. That protects the assembly and makes it hard to apply a lot of sideways or pulling forces on the push plate. I used 1/8″ tubing instead of rod so the glue could bind to the inside of the tube a bit. It also makes it so you can fill the hole with glue, and when you insert the tube the extra just goes up inside the tube instead of squirting out all around. I chose stainless over brass to be more food safe. The springs aren’t stainless though, so they really shouldn’t spend too much time in contact with water.

Once I had assembled and glued the cutters they had to be clamped to keep the springs from tearing them apart before the glue was fully cured. Here you can see a two Moomins clamped in the jaws of a big Jorgensen Clamp. Don’t worry. I was gentle. I managed to get them all done in time for Christmas *phew* but I did end up having to give them away 100% untested. Eek!

The day after Christmas I finally got to take them for a test drive. I had had to make the cutting walls a bit thicker than I would have liked, in order to have enough surface area for the solvent welding. That made it so you have to push down a bit harder than with the comercial cutters, and give a little bit of a twist or your Moomins end up with a paper thin fringe, but ultimately they worked fine. Giving gifts I hadn’t fully tested was a bit hair raising, but it all worked out in the end.

The details were crisp, and after some baking so were the Moomins. Mission accomplished!

Pastry Cutter Progress

January 17, 2013 by Kurt

In this project I only had two disasters. I often keep my projects in sewing boxes because they’re cheap, stack well, have semi clear sides (so you can see what’s in them), and have nice rubber handles. The down side is that they do not have locking lids. That used to worry me, but in all the years I’ve been using them it had never bitten me. Until this year. I was hauling the box out of the trunk of my car when one of the latches caught on my jacket pocket, and dumped everything out into my driveway. In the dark. In the pouring rain. Trying to find tiny clear plastic Moomin facial features in that torrent was no fun.

The other disaster was more burocratic. I have Wednesday nights to do these kinds of projects. It’s my night off from being a Dad and I often spend it at TechShop. Normally booking the laser is pretty easy, but in the run up to Christmas the schedule gets filled up by busy Christmas Elves. This year I forgot to make one of the weekly reservations and was suddenly looking at doing much of the lasering on the day before Christmas Eve. That put a big kink in my gluing schedule.

I used two kinds of glue in this project. Most of the parts are bonded using acrylic solvent welding, and the bond between the metal push rod and the plastic cap/base was done with J-B Weld’s Kwik. I knew that aligning and solvent welding the thin “outline” parts that form the cutter would be a pain, so I laser cut a wooden form that could be placed inside the sections and force them to stay alighted while I applied solvent around the edge with a syringe. Not being plastic I wouldn’t have to worry about them getting glued in place. I also made the alignment pieces have some “cutouts” so those sections would be easier to get in/out and wouldn’t pull solvent away from the plastic pieces via capillary action. The only downside was that these forms left a small amount of black charred-wood-crud on the insides of the cutters, but most of that could be removed by the light sanding I needed to do anyway.

I made Moomintroll’s face out of a lot of tiny separate pieces. I figured that way each Moomin face would be slightly different and that would be nice. What was less nice was the large amount of time spent positioning every eye and eyebrow with tweezers, and the addition time needed for swearing every time my syringe full of solvent knocked one of those tiny little pieces out of place. On the other designs I grouped the pieces together. Snufkin only has his face, pipe, and hat feather. MUCH easier to position and glue. When I suddenly was faced with a week less time to do this gluing I decided to build something to make the gluing go faster. I thought “what if I build some sort of vacuum clamping setup?”

I wanted something that would hold the parts in place as the solvent did its work, and also let me see what was going on in case something got knocked askew. I got a fish tank air pump, reversed the valves so it was sucking instead of blowing, and used that in conjunction with two 1/4″ acrylic sheets and some weather stripping to make a vacuum clamping area. I thought “I can apply the solvent, put the cover on, turn on the pump, and I’ll be able to see everything. This is the sort of desperation that comes from suddenly loosing a week of gluing.

Was the vacuum clamp a success? As it turns out the weather stripping leaked enough to make the clamping action very slight, but it did make a nice clear box with just enough clamping action to hold the facial features in place while I then piled a big heavy chunk of metal on top to provide the final clamping pressure. So all in all I judge the vacuum clamping chamber to have been a “modest success” instead of a complete waste of time.

Gluing and painting. To make the images of the characters on the tops of the cutters I laser etched the designs though the acrylic’s protective paper. Then I used a piece of rubber to squeegee the paint down into the etched areas. Finally once the paint has dried some I peal the paper way. It’s best to do that before the paint has fully dried because the fully cured paint is stronger and pulling the paper away can pull bits of paint out of the design. When the paint is still weak it comes away clean.

Building a Motorized Iris Diaphragm

August 9, 2010 by Kurt

There’s something cool about a mechanical iris. The opening expanding and contracting like magic, the nice radial symmetry. Very fun. Lawrence and I had built the Solar Plotter, and although the output was very nice it has some limitations. Because we were using a large cheap magnifying glass the output lines were pretty fat. I thought it would be cool to have some sort of dynamic line width control, and what would be more steam punk and fun then a magnifying glass with a motorized iris.

The First Build

I started by making a laser cut prototype. I used poster board for the shutters and 1/8″ acrylic sheet for the control rings. The whole thing was pinned together with a few 1/8″ brass pins. The iris worked fine, but there were a few problems. One was that the range of motion for the control levers was not that large, and because it had a sort of snap action it seemed like it would be hard to get fine grain control of the hole’s diameter. Also because there were only 5 shutters the opening was a pentagon, and that might well show up in the final drawing. (If you look closely at the burned wood output of the Solar Plotter you can see little discs stepping along in the ash.) I’d designed it this way to cut down on the number of identical brass sheets I was going to have to eventually hand cut to make the final (longer lasting) shutters.

The Second Build

So I canned that design, and went with a different style of iris. I saw this page about a different iris design. The new design has 12 arc shaped shutters with a pin that sticks out one the bottom at one end and on the top at the other end. I was originally thinking of biting the bullet and snipping these things out of brass sheet, and soldering down short segments of brass rod to act as the pivot and drive pins. As I tried making them it became clear that getting all those parts flat again after all that snipping/soldering was going to be a real pain, and I worried that if the shutters weren’t super flat they’d interfere with each others motion.

All that brass was going to add a lot of weight that I didn’t really want. Of course nothing is more steam punk then a brass motorized iris. But it was time to give up on the complicated hand fabrication of 12 shutters and instead figure out something that I could cut directly on the laser. So I went back to the poster board. The great thing about the new design was that there was very little stress on each of the shutters, so all I really needed was a good way to glue the drive and pivot studs onto the thick paper and I’d be in business. It would have been nice to use very thin plastic sheet for the shutter blades, but I can only cut acrylic on the laser, and that doesn’t seem to come in super thin sheets. (Or at least Tap Plastics doesn’t have it.)

I redesigned the drive plate and bottom pivot plate so they’d accept much larger pins so the pins would have a larger glue area in contact with the paper. I used wooden pins thinking that I trusted Elmer’s Glue and wood bonding much more then some other random glue and acrylic. That was an OK decision, but if I had it all to do again I’d user acrylic pins. All that sanding and champfering of 1/4″ wooden pins was fiddly work and burned more then an hour of time.

The result was a very nice smooth acting iris driven by an RC hobby servo.

Here you can see the gear that goes on the servo motor (center) the drive ring with it’s distinctive drive slots, and the two arc pieces that make up the back wall of the iris.

Here you can see all the shutter blades still in the laser cutter. Only 12 are used, but for this kind of thing it’s always a good idea to make extras. The rounded end it the pivot end, and the other end is the driven end.

Here you can see the basic layout of the drive ring. On the right the servo would turn that gear, and the drive ring rotates in the cup of the back wall. Each of the slots drives the end of one shutter blade around. It’s a bit wonky to have this motion without a nice central pivot point but that’s the nature of be beast.

Here you can see the pin around which each of the shutter blades pivots. I’ve just glued all 12 of them. You can see how I had to sand the ends/edges to get rid of any splinters left from cutting them on the band saw. That was a fiddly business and I highly recommend just using laser cut acrylic discs. You’d only have to sand one face, and then use some sort of glue you trust to do a strong acrylic to paper bond. I’m going to do some glue experiments and figure out what works well for this situation.

Here you can see I’ve flipped over the shutter blades and glued more drive pins onto the other end of each one.

Here you can see all the shutter blades in place on the bottom ring. Each overlaps the other around in a ring. Because the shutters go less then half way around the circle there is never a thickness of more then 6 shutter blades at any one point, and the thickness is always very uniform. The hole in the foreground is where the brass pin will eventually pin the semicircular back wall to the base.

Here you can see me taking the iris for a spin. I put the drive ring over all those pins and gave it a twist. The action is very smooth and controlled. The first layer of the back wall is just set (approximately) in place for this photo.

This is an old servo that I had lying around. It has some little screws that will self tap into those holes to hold it in place.

Here’s a bottom view of the iris with the motor in place. When I mounted the motor I realized there was something wrong! The gear didn’t mesh right. I was off by a whopping 0.075″ Crazy talk. I was very worried because I didn’t know where the problem could have crept in. Had I had more stuff selected when I was nudging elements in Illustrator? What else might be wrong? Then I realized what was wrong. I had flipped the design for the bottom plate so that when it cut out the side away from the laser could be the “top” I’d done this because the holes are always smaller at the bottom side (because of the somewhat conical cutting action of the laser) and I wanted the edge that would be contacting the pins to be as close to the base of the pins as possible. However I had forgotten to flip the motor over, and it was now off center. (accounting for some of the extra space.) Phew!

I flipped it back over and we were good to go.

Here’s the final assembly. There’s still a little bit of play (you can see the gap between the drive ring and the back wall) If I end up cutting a new base ring, I’ll bump that motor bracket in an extra 0.04″ and that should be perfect. It runs nice and smooth. But there’s a bit of a problem. The servo rotates more the 180deg. I had designed the mechanism for 180 of rotation plus some modest amount of slop, but on this motor it seems like it might be an extra 15 deg or more. So much for believing something I read on the internet! The problem is that if the motor can forcibly drive the iris past it’s comfort zone (where the wooden drive pins have bumped up against the outside of their slots in the drive ring) the iris might be damaged, and since we don’t know how good the control will be for the servo we have to assume it will frequently be run to it’s limits in both directions. We really should make sure that the iris is OK in those situations.

The easiest solution would be to make the slots a tiny bit longer. Another solution would be to make the drive gear a little bit smaller and scoot it over some. It’s surprisingly difficult to find exact rotation range number for this model of servo. I guess most RC folks don’t care at all.

The original page in the Internet Archive.

Building a Mechanical Cryptograph

January 8, 2010 by Kurt

This year I’m building an encryption device for my Retrotechnology Society. If you recall last year 6 of my friends/relatives were “antecedently” inducted into a secret society. The only down side being that the society was so secret that I never heard back from anyone. This year I’m going to change all that by providing them with this encryption device, so they can communicate with me (and one another) in cipher. Functionally it’s based on the 1850’s Wheatstone Cryptograph, design wise I’m working on spicing it up a bit.

These images are of the very first prototype. It’s totally rough. I want the basic shape to be the Retrotechology Logo of the 11 toothed gear with an eye in the middle. I’m trying to design the eye so that it “looks around” as you encode/decode messages. You turn the thumbwheel at the upper right, and that drives two gears which turn two rings of letters which are visible though two openings in the face of the device. The eye design is really rough, and various spacings will be adjusted, but this is mostly just a proof of concept to make sure things are actually going to be functional.

These were laser cut out of clear 1/8″ acrylic. Later versions will have to have clearance areas etched into various parts, etc. (to prevent binding) and I’ll have double brass sleeves acting as the bearings. (right now it’s all just jammed in 1/8″ brass rod to hold it together enough to make sure things were working out.) I like how it’s coming out, but I think the aesthetic of the eye design really has to be worked out some more. Still I think it’ll be fun. The final color scheme is going to be black, with white disks inside. I haven’t deiced yet what to do with the back. I could add some cutouts that would expose the gearing a bit. Kind of a skeleton Cryptograph, but that might make it less functional. “keep your fingers and crud out of those holes!” so I haven’t decided. Maybe just a redux of the eye.

I wrote the code to generate the gear profiles in python, and then imported that into illustrator and did all the rest of the work there.

Ok, so then it was time to cut out real versions in the right colored plastics, etc. I used black for the outer casing and “Ivory” for things like the friction wheel, and the lettered discs. In the design the black front/back called for gold and white lines. I thought this was going to be a real pain to do, but I came up with a trick. I used a little rubber squeegee (cut from a Speedball rubber pad) and was able to squeegee gold paint down into all the lines and curves in a jiffy.

This squeegee system reduced the amount of paint used, and really speeds up the whole process. The fills are more uniform too. I don’t want the fills to be perfect because the device is supposed to look sort of old, but the amount of time I was saving painting all that stuff made using the squeegee super important.

Once the splops were on, it was just a couple of strokes to fill every line.

I did have to manually clean out the center hole. It doesn’t matter if there’s a bit of extra gold paint in there since it’s going to have a brass tube pressed in, but we can’t have a thick blob.

If you look closely you’ll also see that I have scraped the outer ring free of most of the gold paint using a paper towel and my fingernail. Then I painted white over the top.

These are the faux Ivory thumb wheels with their special strate knurl, and inlayed arrows. (Done with the same painting technique)

I would have rather had the gears be “ivory” but since they’re not actually visible, I was able to cut them nested inside the other black components, and save a lot of material. I’ve toyed with the idea of making a fancier gear and a “skeleton” version of the Cryptograph, but decided to punt. The dark rings and little “tabs” you see in this are places where I’ve etched extra clearances on the gear/outer wall of the device so that friction wheels can run more smoothly without their knurl making their action feel “notchy.”

I also did some initial tests of burning a deep deep hole in a book. I’ve had people tell me it’s impossible to laser cut books, but you can. You simply have to clamp the pages together so they aren’t as prone to catching fire. I laser cut a clamping jig, and the first cut went well. (Probably 1/4″ easy) but subsequent cuts are bad because the loose paper and forced air cause a lot of burning. (I remove the inner paper, but the walls aren’t as well clamped as they were before and there’s a lot of smoke and mayhem. So I think I’m going to have to change the way I index the rig and flip pages/reclamp after each cut until I get to the needed 5/8″ inch depth. Still here you can see a piece inside a book with only minimal effort. I don’t think I’m going to glue or do anything extra. Just cut the hole so the device can be tucked into the book for safe keeping.

I did this test on a volume of a Funk and Wagnalls New Encyclopedia. (Millions were given away free in super markets back in the day, and it’s hard cover with gold outer trim, so it seemed like a good call.) I really need to find some uniform nice looking old hardbound books that no one wants/cares for. Not sure how I’m going to do that, this thing was $4 at Goodwill. I have no idea why anyone but me would buy such a thing, and that’s a kind of steep price! If I could have snagged 10 volumes of this F&W I might do that, but I’ll have to keep looking.

So this morning I started assembling a final unit. Here’s the parts mostly laid out. They gray part of the eye is supposed to look a bit rough like that. I want the device to seem a bit old, not super crazy snappy new.

Here’s the body assembled with its 11 pins in place. (11 pins! I have to cut and deburr all those by hand, 1/8″ is a bit small to automate on the lathe. (Kill me.) You can also see the super fancy inner pin that drives the middle disc via a square drive shaft that passes though the center of the other gear.

Here’s a closeup of the friction wheel on it’s shaft. You can see it’s inner brass sleeve.

Here you can see the most complicated brass part for the project. Most of the 18 brass pieces are just bits of rod or tube, but the central drive shaft is made up of a brass rod, and square brass tube, and then a custom machined brass tube (the needed to be machined so that it nested to the outside of the square tube and then nested properly inside the 1/4″ brass tube sleeve that is in the center of the upper gear. This shaft lets the bottom gear smoothly drive the lettered disk that is on top of the upper gear.

I was originally going to solder the bits together, but they pressed together so nicely all I really had to do was add a dab of JB Weld to eventually fix the central rod to the inside of the square shaft. This let me adjust things up/down slightly on assembly. Which was nice.

Another device nearing completion.

Here you can see four of the devices, three done, and one just needing to have the bottom paper pealed off. You can also see last years wax seal, and some sealing wax because I’m about to put the devices in their books and package them up for shipment.

The process of cutting the books was a real messy pain. In order to be able to cut a lot of pages at once I had to make this book press which clamps 80 some pages at a time, and lets me do alignment of the cut. You can see the little ‘L’ shaped page alignment guides for positioning the corners of the page.

The problem with cutting so much paper is that you have to go slow, and the compressed air blowing in the machine makes the paper char a lot more then it would for cutting just a few sheets. So there’s a lot of char and mess, and although you only spend maybe 3 mins in the laser doing each cut the time it takes to do a cycle (take out the old paper, clamp up the new paper, and tape things up so various pages don’t fly around while the laser is cutting, etc. eats up a lot of time, and you have to watch the cut like a hawk to make sure it doesn’t go to open flame, or something comes untaped and is flapping around, etc. Nothing I’ve ever done on the laser before was “messy” in this way.

In the end I did manage to cut 6 books before Christmas.

I was lucky to get this 1909 lavishly illustrated 12 volume set of “Adventures in Bookland” which was great because they were all the same size, which simplified alignment. I got all of them (plus some more books) for $5 at the Friends of the Library book sale. So it was a great deal, but I paid a heavy price in terms of guilt. It was horribly Fahrenheit 451 and I felt awful cutting though all the stories I knew like Robinson Caruso, etc. I’ve kept all of the cut out illustrations for possible future use, but my hands have run black with the blood of many an old book and that made me feel bad.

Here is an image of the device in a book. I love the way it looks like the device has burned its way into the book.

Here’s a close up of it in the book. The “top” page here is quite burnt because it was exposed to the most compressed air laser action. Intermediate pages were much less burned looking, but every sixty pages or so theres are a few really burnt ones because that was the next layer I cut. Alignment between the layers isn’t perfect, but it’s decent. You can see I added a faux window reflection that holds still as the eye looks around. I’m really proud of the way this turned out. Now to see if I get an encrypted messages!

Retro Tech Journal

Tales of Creation and Recreation

Category Archives: Project