Neil's Puzzle Building Blog
15Jul/143

Reproducing a puzzle

It's been a little while since I sat down and wrote anything, and it's not for lack of puzzles to write about, but really because I've been so busy in my own shop making puzzles. With IPP looming, I had set myself some lofty goals of designing and producing my first exchange puzzle. That in itself was a lot of work, but I didn't stop there. I went on to make copies of Iwahiro's ODD Puzzle, then a small run of puzzles to give away to new IPP attendees, and then I took on my most ambitious project to date. I remade a Stickman box.

The finished box

The finished box

If you're a regular reader, you'll remember that some time back I won a Stickman Domino Box at auction, which needed a little work to turn it into a fully working Domino Box. Having made the Dominoes, and learned a lot about the box along the way, I had the insane idea that I should make a few copies of the box for other puzzlers. There were only 25 originals made by the Stickman Puzzle Box Company, and it's such a fun design, it really should be played with by more puzzlers. So with insanity on my shoulder, and time slipping away beneath me, I embarked on the most challenging project to date.

A new batch of Dominoes

A new batch of Dominoes

First off, I needed a new set of dominoes, so looking around the growing stockpile of exotic woods that I've purchased, I selected a number of boards, and set about the same process as before, and made a bunch of dominoes. I'll not go into the details, as I've covered them elsewhere, but in total I made up 11 sets of dominoes.


Puzzle Piece Bevel Jig

Puzzle Piece Bevel Jig

The last time I'd made the dominoes, I created the bevel on each edge using the router table, with a 45 degree bit. In fact that's how I've done all the bevels for every project so far. I've never been too happy with that as a process though. The pieces I tend to work with are a little small, and with a high speed rotating blade, it wouldn't take much for a piece to slip, and things to go wrong quickly. So taking some advice from fellow puzzle maker Eric Fuller, I created a new jig. As you can see, it's a 45 degree cradle, where the saw blade pokes out through the centre. This allows me to run the piece to be beveled over the blade pretty quickly, and the amount of the blade sticking out is so small, that it's almost impossible to stick a finger in the blade. Safety first!

A wider shot of the beveling jig

A wider shot of the beveling jig

As you can see from the wider shot, the jig is clamped to my fence, and also clamped to the table, so it won't move anywhere, making this a pretty safe jig. It's also fast. Since I don't need to put a push block onto the piece, and then guide it over the spinning router bit, the whole thing takes much less time.


Drilling Domino pips

Drilling Domino pips

Much like the last time I made dominoes, I had the jig and templates to cut the dots or 'pips' for the dominoes. With 11 sets, or 308 dominoes, or 1848 pips depending on how you want to look at it, there was a lot more work this time round. I'd learned from the last attempt when I ended up with blisters from where the drill rubbed against my fingers, and taped up my fingers at the appropriate points to save any further blistering. Sadly, not everything went to plan.

Disaster

Disaster

Around ¾ of the way through the process I hit a snag. My template drill bit exploded! The bit works by having a collar which is held onto the main body with an angled internal neck. So the collar can't come past that neck as it increases in diameter. The spring pushes it back to the point where the drill bit is covered, and when you push won inside the template the bit is exposed, allowing it to create a pip. Well it seems that since this drill bit made all the original sets of dominoes, it was just rather worn out, and the neck had work to the point where it no longer held the collar in place. With this, it was impossible to finish the dominoes, and I had to scramble to fins a replacement. Fortunately, Rockler sells the kits, and replacement bits ... so by mid week I had a replacement and could finish the pips. Dominoes complete, it was time to turn my attention to the puzzle carcass.

Basic pieces for the carcass

Basic pieces for the carcass

Before starting on making the boxes, I'd been studying the design of the joinery, and how the box had been made for weeks. I'll freely admit that I really didn't understand why Stickman had done things the way he had, but I had a good idea of what I was doing, and having taken lots of measurements, I proceeded to start making sawdust. Having surfaced and thicknessed the boards, I cut the basic side panels, the top and bottom, as well as the corner posts, and top/bottom frames. The wood choice was Katalox for the body and Cocobolo for the frame. The Katalox will darken over time to a dark black almost ebony like colour, but starts off as a warm rich chocolate brown. The contrast with the Cocobolo should be quite striking.

Initial test fit with rubber bands to hold things together

Initial test fit with rubber bands to hold things together

After an afternoon's work, I had all the various grooves cut to both support the panels, and allow the top and front to slide respectively when the box is opened. At this point, I came in from the shop, and announced to my wife that Stickman is a genius. Having not truly understood why the box was created the way it is, and the reasons behind design choices, previously, I now fully understood why each piece existed, and exactly how it interacted. Not only is the construction a work of art, it also requires the minimal number of cuts, and pieces to make everything work perfectly. There is an elegance to the design that I hadn't appreciated. Having spoken with Stickman about it, he agreed. It's only once you've tried to repair someone else's work, or remake it, that you truly appreciate the process they went through to make the final design. And yes, I'm lucky to be able to talk to the original designer, and get his input and feedback as I worked through this project!

Yes, that piece is still attached

Yes, that piece is still attached

With the main body cut and fitting perfectly, I turned my attention to the 'decorative' elements of the puzzle, namely the corner pillars. I'd cut the pieces to the right dimensions but needed to cut out the centre to create corner sections. My first pass on the saw left a rather fun little piece, but after a small adjustment, things were spot on. The joy of working with jigs is that you can make a small adjustment, re-cut, adjust until you're good, then just cut lots of exact replicas of a piece quickly and accurately.


Cocobolo matchsticks

Cocobolo matchsticks

When cutting the corners, I was left with a lot of cocobolo matchsticks. As with any exotics in the shop, these all end up in a scrap bin, and will eventually be used for something else. Nothing gets wasted!


Initial circular holes in the side plates

Initial circular holes in the side plates

The original puzzle is peppered with holes and gaps. They're not just because Stickman wanted to make the puzzle lighter (although in my case, working with Katalox which is a dense, heavy wood it's not a bad thing) but they're integral to solving the puzzle. You need to be able to see what's going on inside, and be able to poke your fingers through the sides to manipulate the dominoes. Stickman did all this work on a CNC machine, where effectively, he clamped the wood down, and let a computer controlling a router bit do all the hard work. I don't have a CNC machine, but I did need to replicate all the holes, and they had to be pretty close to the originals since larger holes would let dominoes escape.

To create the ovals in the sides, I started at the drill press with a Forstner bit and drilled the centres out. I had planned to then take the pieces to the scroll saw, and carefully cut out the ovals. Having tried this, on the first piece, I was entirely unimpressed with the result, and decided not to proceed with that method. Instead I went back to the drill press, and drilled an additional hole either side of the initial central hole, creating something close to an oval. Then I spent several hours with a Dremel to clean up the inside of the hole, and make the ovals smooth and consistent over all the panels. It might seem like a lot of work, but in the end it was a lot less work than my original idea would have been!

Time for a game of \

Time for a game of "Join the Dots"

While I was drilling 'ovals' for the side plates, I also drilled a lot of small holes in the top and bottom to allow me to create the stepped holes I needed. There's no clever way for me to do this without the CNC machines, and I had to do this by hand. Drilling holes at the corners allowed me to insert the blade of the scroll saw, and cutout the holes I needed. By having a hole at each corner, I can easily turn the blade, and get square corners. Before I could start working at the scroll saw though, I needed to have a template to follow, so after a game of "Join the Dots" I had reference lines I could follow.

Inserting the blade through the holes

Inserting the blade through the holes

This shows how the blade is inserted into the middle of the piece, and lets me cut out the hole without having to cut through the sides of the board. It might seem obvious, but there are very few tools in the shop that allow you to make such a cut, and the Scroll Saw is the only one that is close to safe, and accurate enough. The blade I was using cuts a saw kerf that is just 0.01" thick and leaves a very smooth finish, meaning that there was little additional work after cutting. Of course, I still had to take my time, and each plate took around 45 minutes to cut.

Forgive my Scottish knees

Forgive my Scottish knees

With one panel fully cut, there were only 13 more to cut. Over the course of the next couple of days, I slowly worked my way through them (and a good number of fresh blades ... Katalox is dense and dulls blades quickly) before sticking some sand paper to a lollypop stick, and sanding the edges of the holes to soften them, making the piece much nicer to hold.


My first dry fit

My first dry fit

With all the components cut, I could put the pieces together for the first time, and get a feeling for how the final puzzle would look. There was still a lot of work to be done before I could start gluing up the pieces, but this gave me a good feeling for the fit and the look of the finished puzzle.


Pre-finished front and top panels

Pre-finished front and top panels

Given that I would have to glue the puzzle together with the front and top panel in place, I went ahead and pre-finished those pieces. Once they were in place and the corner posts were added, there would be no way to remove them or access the insides to be able to apply finish. The other benefit to pre-finishing the pieces is that the lacquer and wax would help prevent the glue sticking to them, so if there was any glue squeeze-out, it wouldn't lock the puzzle solid. After all the work to get to this stage, that's the last thing I wanted!


Piles of puzzle box pieces

Piles of puzzle box pieces

With the boxes mostly taken care of, the dominoes made, there is one additional element needed. There are five additional pieces in various shapes that go into the box to make things more troublesome for the solver. I decided to make these from Redheart and Purpleheart. Each piece is a solid section of wood, and I cut the notches out of the piece. This gives me a nice continuous grain, and a stronger piece. Not to mention that there's less work for me to cut out a piece than having to cut twice as many pieces, and then re-glue them later. As with the dominoes, these pieces are passed over the bevel jig, to soften the edges, however there is a small problem. Given the internal edges it's not possible to get the blade into them. This meant that each internal bevel is cut by hand with a chisel. So lots of extra work from my side, but I end up with a clean consistent bevel around the pieces.


Internal blocks glued in place, ready for the final glueup.

Internal blocks glued in place, ready for the final glueup.

On the front panel, there are two blocks which are glued to the inside which are the final part of the locking mechanism. A gap needs to be freed below these blocks to allow the front panel to slide, and then the top to slide. Each piece needs to be added at the right time, or you end up being unable to glue the piece in place, and end up with firewood rather than a puzzle.

By this point it seemed like I'd spent weeks and still had nothing more than a pile of pieces to show for my work. The glueup for the box is rather complicated. Unlike most other projects I'd tackled, the box can't be glued up, then assembled. The glueup is the assembly. The way things work is that the front panel, and top panel both have 'wings' which travel in the grooves of the side panels. In each panel, a section of the wings are removed to allow them to pass the pillars on the corners and the trim on the bottom. So when I mentioned earlier that Stickman is a genius, this is why. Those pieces which look like trim and decoration are actually integral parts of the mechanism. What is means is that for the glueup, the top and front are inserted, then the pillars are glued in place. This prevents the top panel from being removed again, so I had to make sure that everything was in the right place.

Seven Domino Boxes Standing in a row

Seven Domino Boxes Standing in a row

So with the blocks in place, the pillars glued on, and a check that everything still moved, I could glue the top and bottom frames in place, finishing the glueup, and being the moment of truth as to whether I had a puzzle box reproduction, or some very expensive (and frustrating) firewood.

While the glue on the frames was drying, I clamped several boxes together back to back and clamped them together. This meant that each box was providing a gluing jig for the one next to it, giving me good clamping pressure across the entire joint, and ensuring that all the boxes were identical.

This is probably the most challenging assembly I've attempted, and I'm pleased to say that it all went very well. I've clearly learned a lot over the last couple of years, and having three times as many clamps today as I had when I started certainly doesn't hurt. It really is true, you never can have enough clamps!


The difference is marked after applying lacquer

The difference is marked after applying lacquer

The following night, once the glue was dry, I could finally test the mechanism on each box, and confirm that it worked as intended. You have no idea how much of a relief it was that each box slid smoothly and opened as expected... until the last box. The lid didn't come out past the front pillars! Fortunately, I have some sharp chisels, and I took a few thousands off the top of each of the front pillars, and the lid slid past. The amount I had to remove was so small, it's unlikely that even I could find the box I had to shave, but it was enough to stop things working. Fortunately, it wasn't a difficult fix.

The last steps were to finish the puzzles. As you can see the boxes at the back where the first coat of lacquer has been applied really show the true colours of the wood, unlike the boxes at the front which look drab and liefeless without the finish.

Sets of dominoes all finished

Sets of dominoes all finished

The final steps are to triple buff the puzzles and dominoes, to really make them shine.

Stickman was kind enough to send me the original files for the puzzle booklet, so with a couple of updates, I have been able to create a new booklet which matches the reproduced puzzles. (Ed: Have I mentioned before how nice a guy he is?)


Stickboy

Stickboy

While I was working on the puzzle, I was posting images of the work in progress. Seems that a few of my puzzling friends have suggested that I could be "Stickboy". I might have taken that idea and run with it. Stickman seemed to like the idea, and agreed to let me use the moniker and a modified version of his logo. I also took his own logo, and created a much higher resolution version for use on Stickman's next project. After all, I was cleaning things up for my own use. Seemed silly not to let Stickman benefit from my work.


The finished box

The finished box

After that, it's a simple case of putting the dominoes inside the carcass, mixing things up and passing them to a new owner. It's been an interesting project, and one I'm glad that I tackled. The question is which box do I tackle next?



2Jun/145

ODD Puzzle

Way back in 2011, when I was at a puzzle party here in the Bay Area, I had a play with a clever little puzzle, by Iwahiro called the ODD Puzzle. A simple three-piece packing puzzle, that was anything but simple. It's taken me three years, but I now have a copy of my own .... and I had to make it myself!

Odd Puzzle by Hirokazu Iwasawa (a.k.a. Iwahiro), made by TheJuggler

Odd Puzzle by Hirokazu Iwasawa (a.k.a. Iwahiro), made by TheJuggler

Back when I first saw the puzzle, the only copy available was a version by Philos. It was made from Beech and measured 4.7" x 4.7" x 3". The fairly plain appearance hid a superb puzzle, that was the grand prize winner at IPP28. Now it's mostly sold out everywhere you look, and no real indication of when or if it will be available again. The only place I've been able to find it is at Puzzle Sport in Germany.

I played around with the copy at Stan's and really liked the puzzle. It was clever and really needed you to spend time playing with the pieces in the box to understand just how they interacted with each other, and what movements are possible. I was unfortunate that the puzzle was sitting in its solved state when I picked it up, and it was more of a challenge to remove the pieces for me. If I'd not seen the solution I think I would have required a lot more time to understand how it would be possible to get the pieces in there. I had mostly forgotten about this puzzle which if I'm honest is rather a shame.

The Facebook picture that started all this.

The facebook picture that started all this.

Fast forward three years, and I spotted a Craftsman copy of the puzzle made by Iwahiro on one of my friend's Facebook pages as his profile picture. (ed: There's a funny story there, but I'll come back to that.) Seeing that picture spurred me into hunting for a copy, and having failed, I asked a couple of my puzzle friends if anyone had a copy of the puzzle. Amusingly, the closest (geographically) to me had a copy of the puzzle, and agreed to lend it to me. A couple of days later, we met up and exchanged a few puzzles, had the obligatory play with a few new additions to each of our collections, and he left me with a copy of the ODD Puzzle, as well as a puzzle from his collection where a couple of glue joints had broken, that he asked me to fix for him.

Fairly quickly I took a host of measurements from the reference copy, and set about making a copy of my own. I really liked the look of the craftsman version with the 'core' having a different wood to the outer parts of the pieces so looking through my stock, I picked out some Lacewood and Yellowheart to make the pieces. Rather than describe the process in detail, the images below should show the highlights. (As ever, click for larger versions.)

Initial glueup of the pieces

Initial glueup of the pieces

Basic blocks cut

Basic blocks cut


ODD shapes, 'O' still too thick

ODD shapes, 'O' still too thick

Boxes cut, and ready for glueup.  Base pre-finished.

Boxes cut, and ready for glueup. Base pre-finished.


Box glued up, unsanded and still no finish yet

Box glued up, unsanded and still no finish yet

Finished!

Finished!


Of course as you can see, when you're making one, it's not a lot more work to make two, or three, so I made three copies. When I was posting pictures, on my own Facebook pages during the build, I was contacted about buying copies, so I'm glad I made extra. One interesting build fact is that I pre-finished the bases before I glued up the boxes. The reason being that I was going to be unable to finish the inside of the base once the box was glued together. One benefit to pre-finishing is that the glue won't stick to the finish, so the base will be truly floating. I also pre-finished the inside walls of the box before I glued on the strips that create the gap. Again, I wasn't going to be able to get a brush inside the box to finish it to the level I wanted, so I finished it first, leaving clear areas where I was gluing up the pieces so that the glue would stick.

Iwahiro's version, next to my copy

Iwahiro's version, next to my copy

As a quick comparison, the pieces for both copies are identical in size, and can be interchanged between the boxes. In Iwahiro's version he has mitred joints in the box with slip feathers for strength. I used a shoulder joint on my box and didn't add splines (a.k.a. slip feathers). I did however change the way the top of the box is made, extending my slats all the way through the sides, making that joint stronger, unlike Iwahiro's version where the slats are entirely inside the box. There's not real difference, just a choice I made.

A look at the bottom of the puzzle made from Chakte Viga

A look at the bottom of the puzzle made from Chakte Viga

My the boxes in my copies are made from Katalox with a Chakte Viga base. I think the contrast between the dark, almost black sides with the bright orange base looks great, but then I wouldn't have done it if I hadn't.

This was the first time I'd used the Beall Wood Buff system since I bought it. Fellow puzzle maker John Devost uses it on most of the puzzles he makes, and having used it now, I can see why. The finish achieved is incredible. It's a high gloss shine, which really makes the wood look great, and gives it an amazingly smooth feel. It also means that the pieces in the ODD puzzle slide against each other with virtually no friction. That means that the slick plastic sheet added in the bottom of Iwahiro's version isn't needed in the copies I made.

One of my friends commented that the only thing missing from my copy was the slip feathers on Iwahiro's box. I joked that I'd considered adding Holly slip feathers, but in the end didn't do so. (Now I didn't mention at the time that I didn't have a jig built that would allow me to create slip feathers in a box) All I can say to that now is ...

Different wood choices, and slip feathers added

Different wood choices, and slip feathers added

They're not quite finished yet, (and I'll update this when they are) but I've added the slip feathers, purely for the aesthetic of them, and rather than have them flat, I put them at an angle just to add detail.

I mentioned earlier that there was a funny story related to me making the puzzles. As I mentioned, I was prompted to make these having seen the picture on a friend's Facebook. After posting pictures of the copy I'd made, the friend who had the picture sent me a message asking if I could send him measurements of the puzzle, so he could make his own. He didn't have a copy, and seeing me making my copies kicked him into deciding he should make his own copy, since he'd been meaning to do it for a while. So, I was prompted into making a copy having seen his photo, which prompted him ... Well it amused me at least!



28Nov/116

Unhappy Childhood

No, this isn't a post about how I had a hard life growing up, or anything of that nature. I had a pretty happy childhood as it happens. Most of you will know already, that this is one of Stewart Coffin's puzzle designs, #41 in his numbering system, consisting of 10 pieces, made from 5 cubes each, which come together to form a 5x5x2 rectangle with a checkerboard pattern.

Unhappy Childhood, boxed in the unique checkerboard solution.

Unhappy Childhood, boxed in the unique checkerboard solution.

This particular copy was made by me and is made from Rosewood and Maple, with a Myrtle Burl box. It measures 3.7" x 3.7" x 1.5" for the pieces, and 4.25" x 4.25" x 1.7" in the box.

This is a pretty tough puzzle to solve, as there is only one solution where you end up with the checkerboard pattern on both bottom and top as you can see in the picture above. There are however 2,408 possible solutions if you ignore the checkerboard. So no shortage of ways to get a 5x5x2 solution! (Stewart Coffin reports that "a computer analysis by Beeler, these pieces pack into a 5 x 5 x 2 box 19,264 different ways", however Burr Tools shows just 2,408)

The following is a look at the creation of this puzzle. Hope you enjoy!

This is one of the puzzle designs that I had been looking at making for a while, since it seems no-one has made any in some time, and I don't have one in my collection. Really that's where this all started, looking to add a new puzzle to my collection, and having spent (far) too much on puzzle already this year, what better way than to make it myself.

It all starts with cubes

It all starts with cubes

So the puzzles that I'm making currently are all cube based, and that's where it all starts. 50 wooden cubes, 25 Rosewood, and 25 Maple is the starting point for the UC. The darker tops on some of the Maple cubes at the bottom of the picture is actually the natural wood. Since I love the look of wood, I'm not selectively removing pieces which don't look perfect. After all each puzzle is unique given the grain and natural colour of the wood, which is something I love. When I put the pieces together, I'll orient the pieces so that very little of this is visible, because I'm really aiming for the contrast between the two woods in this puzzle. If the couple I've made, only one has this distinctive colouring on some of the pieces.

50 Cubes to be Bevelled

50 Cubes to be Bevelled

One row done, 9 more still to do!

One row done, 9 more still to do!

Half way there!

Half way there!

Some time later, and I'm done!

Some time later, and I'm done!

This is one of the most time consuming parts of the process (currently). I have to take all 50 cubes, and put a very small bevel onto each edge of the cube. All in all it takes between 1.5-2 hours with my current method. There's been a fair old discussion in one of the puzzling forums about beveling cubes, so I'm sure I can cut this down significantly, but that's going to need a new jig, and some more tools in the shop so for now I'm stuck with what I have.

If you're interested, the checkerboard piece of wood in the pictures isn't some sort of template, it's actually what will become the base of the box that the puzzle sits in. I just happened to be working on it at the same time, hence it ended up in the pictures.

Next up I made the 10 pieces of the puzzle from those 50 cubies, and as it happens I don't have any pics of the process. I'll need to take a few from the next one I make and update this at a later point. Anyway with that done, I turned my attention to the box. I now had dimensions for the box, based on the final size of the pieces, so I took the burl I was using to the saw, and cut it to the right lengths for the box, and created a dado in the edges of two sides, to allow me to get a stronger joint for the corners.

The outside edges of the box

The outside edges of the box

Despite the very small contact area, wood glues are remarkably strong, and will hold the frame together with no issues. In fact, to take it apart would probably break the wood, before the glue would let go. Using blue tape, I tape the corners, (no clamping required) and that will hold the box well enough for the glue to set. I do a quick check to make sure that the corners are square, and leave it to dry, while I turn my attention to the base.

The unfinished base, with the sides sitting on top

The unfinished base, with the sides sitting on top

As you can see, the base is unfinished. The pencil marks were to allow me to line up each of the strips for gluing everything together. As you can see I still have some sanding to do, since there's glue and all sorts on the base. Thanks to the random oscillating hand sander I got for my birthday, it will make short work of that!

A quick dry fit before gluing the box together.

A quick dry fit before gluing the box together.

With the sanding done, I have a quick dry fit with the pieces in place to make sure everything fits as expected before gluing the base in place. Note at this point, Ive sanded the inside of the box to its final point, as it will be pretty touch to get into the corners once it's all glued together, so best do that before the final glueup.

Box glueup from the top

Box glueup from the top


Box glueup from the bottom

Box glueup from the bottom

It's probably worth pointing out at this stage, that I've spent around 3-4 hours making this box. Given that I decided I wanted a checkered base, that meant cutting thin, equally sized strips, gluing them together, then cutting them into strips once dry, flipping the strips to create the checkerboard, and re-gluing, then sanding, etc etc. All in all probably the most labor intensive part of the puzzle build, but hopefully worth it!

Finishing.  First coat of thinned lacquer applied

Finishing. First coat of thinned lacquer applied

With all the individual pieces ready, it's time to look at finishing the puzzle. The box was all sanded on the outside, and it's looking pretty good. I start off by applying a coat of thinned lacquer to all the pieces. It's 1 part lacquer, 2 parts thinner that I'm using. It gives a very thin coat, but does the job or really making the grain pop. If you compare this to the pictures of the dry fir you'll see what I mean.

First Coat of Watco Satin Wax.

First Coat of Watco Satin Wax.

Second Coat of Watco Satin Wax.

Second Coat of Watco Satin Wax.

Once that's dry, the puzzle gets two coats of wax. I'm using a liquid wax, Watco Satin Wax to do the job. I leave the wax for around 5-10 minutes, then wipe off any excess with a rag. This is building up a nice finish on the pieces, but there's still one more step to complete the process. That's a final buffing with some Renaissance Wax.

The pieces, next to the finished box

The pieces, next to the finished box

A final view from the side, showing the effect of the checkered base.  The myrtle burl box almost looks like its floating.

A final view from the side, showing the effect of the checkered base. The myrtle burl box almost looks like its floating.

The final puzzle ready to be played with!

So there you have it. I hope you enjoyed the build as much as I enjoyed making it.



19Sep/110

Hadrian’s Box

a DIY Puzzle Box

Some time back, I purchased a couple of DIY puzzle boxes from BH Pen & Laser Crafts which are kits based on Bruce Viney's designs.

Bruce makes a large number of puzzle box designs, and supplies plans for you to cut your own pieces. He uses an interesting measuring system based on the thickness of the wood, so you can use any stock you have as long as you mark everything up based on thickness.

Rather than cutting all the pieces myself, I purchased the pre-cut kits to make my life easier. I didn't really feel like spending days cutting all the parts myself, not to mention that some of the pieces are pretty small, making for some challenging cuts.

The video shows the full process from start to finish, using time-lapse. Overall, it took around 2.5 hours actually working on the kit. Including time for the glue to dry, it was around 5 hours.

Overall, it's a good kit, despite the small issue I had. Since I'll be receiving some replacement parts to fix that issue, I really can't complain. If you're thinking about getting one of these, I highly recommend the kit as it's both well made, and Bruce's instructions are pretty easy to follow. There's minimal sanding needed, so pretty much anyone should be able to build on of these, and get pretty good results at the end.



18Sep/111

Tier Box by Eric Fuller

Tier Box is a Japanese style sliding panel puzzle box with a few unique touches, designed and made by Eric Fuller back in September 2009. The 18th marks its two year birthday, so I though it appropriate to add this review today.

The box measures 3.2" cubed and is made from Quartersawn Bubinga for the outer panels, and Quartersawn Paduak for the internal panels. Along with that there's a few magnets and some metal pins thrown in for good measure. 14 moves are required to open the box to reveal the space inside, and the same again to close it. Despite opening it fairly quickly, I must confess, it took me many more than 14 moves to close this one back up!

Eric has this to say about the box:

I am very happy with the results of this, my latest puzzle box. The design originates from a sketch I made in Chicago sometime during IPP23. It combines several ideas I have been wanting to implement in a sliding panel puzzle box. The solution requires 14 moves, but those moves are anything but straightforward and are at times downright devious. I had the pleasure of watching many puzzlers attempt to solve it during the course of IPP29, so I can say that difficulty wise it's a nice 15 minute solve for most puzzlers, with several ah-ha's to spice things up. Fully understanding the interactions between all the panels will likely take quite a bit longer.

Tier Box in its closed state

Tier Box in its closed state

There were only 34 copies of the box made, so I have to once again thank Derek for lending me his copy to puzzle over. It's a fun box, and very solidly built. As Eric notes himself, fully understanding the interactions of all the panels certainly does take some time. I was able to open the box without too much trouble, finding it a fairly simple progression from one step to the next. Closing however was not the same story. I probably spent around 5 minutes opening the box, and well over 20 closing it again. At one point I thought I was going to have to give it back to Derek open as it didn't look like I could figure out how to close it!

So from that experience it's a challenging little box. The panels interact in interesting ways with each other, and the only way to truly say you've solved it is in understanding all the interactions. Despite the pins being around 1/16", they really do get in the way!

Tier Box opened, showing the bright Paduak insides

Tier Box opened, showing the bright Paduak insides

One of the beautiful things about the choice of wood here is that the internals of the box being made from Paduak, are protected from UV, so have retained their beautiful Orange/Red colour which will normally fade to a dark brown if exposed to the sun. It's a nice touch to have this colour screaming at you when working on the box.

My only criticism with the mechanism is that the thin sliding panels used in the internals of the box are fairly tight. While this is normally a good thing in a puzzle box, meaning the panels don't rattle around of their own accord, I found that this worked against me when trying to close the box, as my fingers couldn't push one of the internal panels far enough to slide it to where it needed to be through the small gap left when the outer panels were positioned in the correct locations. In the end, I had to get a small tool to help.

Overall, a superb box, that adds a few surprises to a standard sliding box, and creates a satisfying puzzle.



12Sep/1110

Coffin Involute

My first home built Puzzle

After the successes with creating both Square Sticks and Cubes, I had to go do something with them; and see if I could create a puzzle. I decided to make some of Stewart Coffin's designs, and having been in touch with him, he very graciously gave me permission to try to recreate any of his designs, and encouraged me to do so. With that endorsement, I was off and running. Well, almost!

I had to work out which puzzle I was going to create. There's so many to choose from that it's not an easy decision. In the end, I decided to create something that I didn't already own, so I'd be adding to my collection if it turned out to be any good. So I settled on a copy of STC #214, the Involute puzzle. The Involute is the third in a series of puzzles from Stewart Coffin, each an improvement over the predecessor.

My first attempt - Involute by Stewart Coffin

My first attempt - Involute by Stewart Coffin

The first was Convolution, a 4x4 interlocking cube which requires a rotation in the solution. Due to the rotation, some material needs to be removed from one of the cubes in the solution (if you have a tight fit) to allow the rotation to happen. You can read my review here. Stewart Coffin notes that given the rotation, and the nature of cubes (which don't like to be rotated when hard against one another), that this design could be improved. In his book "The Puzzling World of Polyhedral Dissections", he leaves it to the reader to see if they can find a solution to this problem.

At the same time, Stewart Coffin had already solved the problem, and created STC #198, Involution. Again a 4x4 cube with a rotation required in the solution, but this time because of the design of the dissection, no material needs to be removed from the pieces to allow the rotation. I'll not give away how this is done, as it would spoil the puzzle, but I will say it's a simple and clever solution! I was able to play with one of Scott Peterson's copies that he had made on my recent visit to see Scott, so I can say I've solved both the Convolution and Involution puzzles at this point.

The third in the series is STC #214, Involute. This is the final puzzle in the series, and is again an improvement over the Involution and Convolution. Again there is a rotation required in the puzzle, and again, no material needs to be removed for the rotation to take place. There's an extra trick in this puzzle, that I'll get to in a bit which makes it just that bit more devious.

All three puzzles in the series look identical from the outside, each having the same cross pattern on all six faces, so without knowing which puzzle you have in your hand, it could easily be any one of the three. Have I mentioned that this Coffin is a devious bloke?

Thanks to Allard and Kevin who both reviewed their copies of the Involute puzzle, I was able to model the pieces in burr tools, and from that create myself a parts list and a gluing diagram to be able to build the puzzle.

My piece and gluing diagrams for the Involute.

My piece and gluing diagrams for the Involute.

Given that it took several hours to create the diagrams, including the time to create the model in burr tools and so on, I'm not going to give you the whole thing. Not to mention it would spoil how to solve the puzzle (or would it - I'll come back to that thought). But the image above gives you an idea of what I created.

All the pieces needed to build the Involute

All the pieces needed to build the Involute

With the design in hand, I went off to the saw, and using the crosscut sled and my stops, I cut all the necessary cubes to make the puzzle. There's quite an array of pieces there when you see them all sitting together. Also in the picture is one jig I hadn't talked about previously. This is my cube gluing jig. It's not overly complicated, just three pieces of MDF cut and glued together to hold a 4x4 cube cut to my 3/4" stick size which has all edges at 90 degrees, and has been waxed to prevent any glue from sticking to it. I also have three 'end panels' which will distribute the clamping pressure evenly across all the blocks so as not to twist the blocks while the glue dries.

At this point I made something of a realisation. Sitting looking at this array of blocks, and my gluing diagram, gluing up one of these puzzles is hugely complicated. You're working in three dimensions gluing any number of pieces together, all of which needs to be accurate, and with no glue squeeze-out. If you thought Ikea furniture plans were Convoluted, then this is much more challenging!

Dry fit in the gluing jig  - bottom layer.

Dry fit in the gluing jig - bottom layer.

Dry fit in the gluing jig - three layers.

Dry fit in the gluing jig - three layers.

Dry fit in the gluing jig - full puzzle.

Dry fit in the gluing jig - full puzzle.

Dry fit in the gluing jig with clamp blocks.

Dry fit in the gluing jig with clamp blocks.

Next up I placed all the pieces into the gluing jig, to match my plans. This serves a couple of purposes. Initially, it shows me how good the fit is, and also verified that my plans were correct (at least in as much that I had the correct number of pieces). The other benefit to the dry fit is that it allows me to select which pieces I want to put where in the puzzle. Looking at the grain in the wood, I can select the 'nicest' grain to be on the outside of the puzzle, or look at creating grain patterns by selecting pieces carefully from the pile. Given that this was a first ever attempt, I wasn't too concerned with the grain pattern, but I didn't entirely ignore it either.

Since this was the first glueup I'd be doing, I decided to go with gluing up two layers at a time. This meant that I didn't have to work quite as quickly to get the clamps on the jig to ensure that tight fit I was going for. Fortunately, the way the pieces go together, there is a flat surface after every second layer, which was ideal as a stopping point. I also have a smaller glue bottle, where I've decanted some of the glue from my big bottle. This small bottle has a fine nose, and is much easier to work with that the full sized bottle. Given the small amount of glue I'd need for each piece, this is the only way to work.

Involute in layers ready for the glueup

Involute in layers ready for the glueup

With all the pieces separated into layers, I was as ready as I was ever going to be to start putting this together into a puzzle. Fingers crossed!

First layer glued in place.

First layer glued in place.

Two layers glued up and clamped.

Two layers glued up and clamped.

Working reasonably quickly, I glued up the first two layers, and thanks to tips from Scott Peterson, I managed to do so with little to no glue squeeze-out. That's pretty important since any glue squeeze-out will glue blocks together that shouldn't be, making the puzzle unsolvable. You'll notice the fairly large block of wood on the top of the gluing jig in the photo on the right. That's because I only have two layers build at this point, so the puzzle is half way inside the side plates. I needed to add some height to be able to clamp the puzzle effectively.

Glueup completed, clamped and waiting nervously.

Glueup completed, clamped and waiting nervously.

After the glue had set, I came back and added the remaining two layers, building on the two I already had. This time, you can see that the puzzle fills all the space, and there are no extra spacers required. I then had to wait a few hours for the glue to dry properly before I could take the clamps off, and see whether I had created a puzzle or a paperweight.

They may have a been a few of the longest hours I have experienced in a long time. My fiancée was about ready to kill me, as I wanted to go take the clamps off and see what I had, she kept telling me to leave it alone. I was like a kid on Christmas morning waiting to see what presents I had. I could barely sit still! When things had been left for long enough, I was finally allowed to go take the clamps off and see what I had.

I should note at this point, that I have never solved an Involute puzzle prior to making this one. Given that rotations are required in the solution, Burr Tools can show that there is a solution, but it can't animate the assembly for you (or in my case the dissassembly), so I have no idea how to take the puzzle apart. I'm now in new territory, and given that I don't know how to take things apart, or whether the pieces are glued together correctly, and not glued to one another I know this is going to be interesting!

Since I know where the key piece is, I can remove that fairly easily, but then spend the next ten minutes pushing and pulling on various pieces hoping that something else will move in the puzzle. I can see that there is movement in the pieces, so at least it's not all glued together, but I am having real problems in finding the second move. The pictures that follow were taken by my fiancée, so are unedited as I make progress. That grin on my face is real!

The moment where I figure out move 2.

The moment where I figure out move 2.

Now what?

Now what?

Seeing the individual pieces of the Involute for the first time.

Seeing the individual pieces of the Involute for the first time.

As I mentioned, I've never solved the Involute before, so I had no idea how the puzzle was supposed to come apart. The key piece in the puzzle is really well hidden, and without knowing where it was I would have struggled to start, especially not knowing if the puzzle was entirely glued together at this stage. The second move is also very clever. One thing that Stewart Coffin regularly has in his designs is pieces which are created so that the average person will hold the puzzle in such a way that you will be holding the piece you need to move, effectively pushing the puzzle closed and preventing it from being opened. The Involute is no different, and has this very same trick to allow move two. The look on my face when what looked like half of the puzzle slid to the side perfectly must have been quite the picture. I think for me not only was I solving a puzzle for the first time, which always brings a smile to my face, but also it was a puzzle I had built, and seeing it work the way it is supposed to is an ever bigger achievement.

I took the puzzle fully apart, and was left with the eight individual pieces sitting on my sofa, with a huge grin on my face. I then realised that I had absolutely no idea how to put the whole thing back together! In my excitement of taking the puzzle apart, I wasn't paying any attention to how the pieces were coming apart! I then spent the next 15 minutes with my gluing diagram trying to put the puzzle back together. Remember I mentioned that having the full diagram may not help that much! I did get there, and the smile on my face seeing the puzzle back together was truly from ear to ear.

Closeup of the fit between pieces

Closeup of the fit between pieces

The rotation in the puzzle works perfectly, and I haven't removed any material from the rotational piece to make that move easier. The fit of the pieces is superb. It's difficult to tell individual pieces apart as you can see from the closeup above. This makes finding how the pieces come apart even more difficult that if the pieces fitted loosely together as there is no movement between the pieces. In case you're wondering, that tiny gap that looks like there's a chunk taken out of one of the pieces isn't tear-out as a result of a poor cut, but was some natural holes in the walnut. It's also worth noting here that there is no sanding done on any of the pieces, these are all straight off the saw. Many people in the puzzle community have noted that sanding reduces the accuracy of the pieces, and that a good clean cut can have every bit as good a finish as a sanded piece, perhaps better, since sanding is effectively scratching the surface.

Yes, I made a second to prove it wasn't a fluke!

Yes, I made a second to prove it wasn't a fluke!

To prove that it wasn't just a fluke and this was a one-off, I went off and created a second copy of the Involute. So what you're seeing here isn't some clever photography, but the two copies side by side.

Taking one of the cubes apart

Taking one of the cubes apart

And just to show that it works, there's a partially assembled version next to the fully solved cube.

I was really happy with the results. Over the weekend I produced two copies of the involute puzzle, and both have a very snug fit, and I'd be happy to add these to my collection. In case you're wondering, they're made from walnut with redwood corners. And not to sound like an American advert trying to get you to place an order for something you didn't want ...

But that's not all!

There's another of Stewart Coffin's designs that I've wanted to play with for a while. That's his "Half Hour Puzzle", STC #29. So I drew up the diagram for that, and made one of those too! The brilliant thing about the half hour puzzle is that even though Stewart coffin designed it to only have the cube solution, there are hundreds of possible solution shapes that can be made with the pieces. I've created a burr tools file with many of the solution shapes, so if you're interested in a copy of the file, just let me know.

STC #29 - Half Hour Puzzle

STC #29 - Half Hour Puzzle

Three Coffin's

Three Coffin's

So there you have it. Three puzzles in one weekend, all which I am very proud of, and is the start of hopefully great things. As Allard has put it, "One day there'll be a couple of us around who can say that we had one of the very puzzles created by someone the whole puzzling community now knows as the Juggler-guy! :-)" Maybe ... one day.