Joint No. 3

This joint, man, what can I say. Every time I try one of these it is a challenge, seemingly just the right amount of challenge. What looks simple ends up being complex, the complex is simple. What do I want out of doing these exercises? Certainly I’m not going to run off to Japan and apprentice as a temple carpenter. These joints are often quite a bit more complex than what I use in my shoji work. Maybe one day I’ll build a timberframe, I know I’d love to, but that’s not it either.

Its something about the process. Its the same reason you might learn to play Go, the love of the challenge even when playing an opponent vastly more skilled. It doesn’t matter that you will probably lose, the proper attitude is almost more important than the skill.

Every once in a while, as today, I am cutting a joint and get that feeling of right attitude. I stand straighter, my hand holds the saw with strength and suppleness, and I am focused but not focused.


Brought to you this week by the inimitable Jason of My Peculiar Nature:

The layout looks so simple, does it not? I prepared the wedged key material as well as the dovetailed tenon before starting the cutting.


Lilly was asking a thousand questions a minute. Part of me really wants to focus, but how cool is it that a four year old wants an explanation about every single tool that I use and gets to see it in use? When I think about the different paradigm that she is growing up with, the access to unlimited information, it makes my head spin wondering what she will be capable of. First and foremost though is nurturing the natural curiosity that all children have before it is beaten out of them in one way or the other.



I used my brace and bit to lay waste to the mortise. I wasn’t really satisfied with using the bit because I always manage to over cut my line a little bit. Perhaps I should have just made the mortise slightly over sized.


See? The gods of joinery, as Sebastian would say, are frowning when they see a mortise cheek like that. I pared a bit too much off the left front corner of the mortise, and left too much on the right side. For a minute I thought I had botched the layout, and would have to start over to have my tenons properly align.


The mortise for the stub tenon was 1″ in width, the neck of the dovetail 3/4″. I cut the mortise to a depth 1/16″ More than both the stub tenon and dovetail, so I  didn’t want to directly scribe the dovetail from the mortise cut to accept it. So, starting with the layout on the, what the hell do you call this part, removable tenon? I set my bevel gauge to the layout and used it to pare the 3/4″ mortise cheeks to the taper of the dovetail.


The keyed wedge! My god, what the hell was I thinking about all week when I presumed this joint would lay down before me and assemble itself. Apparently that thought did not extend in a constructive way to trigonometry, because I got to this stage and started scratching my head. I cut the material for the key at 1″ of width, square sides. I basically just started using my highly calibrated eyeball. Can you figure out how I did this? I feel like I need to cut another dozen of these keys (though fear not, there are more joints coming in the future that will use shachi-sen). I actually saw an excellent joint that Chris Hall used in a pentagram stool to lock mitered frame corners together:


By a bit of marking I was able to determine the angles for the parallelogram sides. I added another wedge because its twice as much fun!


The layout for the tapering to the shachi-sen started on the dovetailed tenon, coming in 1/16″ from parallel on either side. The sachi-sen remained consistent in thickness, so I measured the depth inward on the bottom to match the top.


Once again the bevel gauge transfers the taper to the mortise walls. I realized that I sould have marked the top width of the wedge to be smaller than the width of my wedge material. I decided to compensate by paring inside of my lines a bit.


The parallelogram angles for the sides was planed in by eye, and then the ends were planed square to the length on a shooting board. I used the same set of my bevel gauge to mark the taper onto the wedge and planed it with a block plane. Only by sheer dumb luck did I manage to plane the taper on the right end of each wedge, the slant of the sides are opposites of each other.


Paring the slot for the shachi-sen in the beam was straight forward enough, checking the squareness of the non-tapered cheek with the bevel gauge set to square.


Assembly! The dovetail slots in with a few light taps of the genno.


Alas! I left the dovetailed-tenon too wide, it sticks up above the rail. It would be nice if my joints always went together correctly the first time, but then they are designed to come apart.


The second try at assembly went better. I had cut my wedges 1/8″ shorter than the depth of the mortise. I thought I would have a bit left sticking up to saw off but ended up driving them all of the way in. The shachi-sen does really pull this joint together! It created a bit of a gap between the post and the dovetail shoulders, but the joint is really solid. Beautiful really, I just wish I had a better grasp on cutting the sachi-sen.

Hatagane for Paneling

I started on the dimensioning of lumber for my fuigo, but realized that I don’t have clamps that can accommodate the width of the paneling I’m laminating. I’ve covered the clamps I make in a previous post.


How about starting with a little sharpening. This is the ‘V’ cutter for my threadbox. These threadboxes that Highland woodworking and Woodcraft are selling are pretty terrible in terms of steel quality, but they do work. What really helps is a small jig to hold the tiny cutter. Something else easy to make and it saves a lot of frustration.


I had a definitive lack of sawn lumber that can take thread cutting well. The standard is probably hard maple or beech, but some local mystery lumber from down by the creek was the best I could do. I brought it inside a couple years ago, so its probably fairly seasoned. It’s actually very nice hardwood, tight grained and not ring-porous. Not terribly straight grained, so I split it with my froe at about eight inches in length. Ode to the froe, what an awesome tool. I split out sections of sound lumber that were then worked down to a rough 1/2″ square with a hatchet, and further refined with marking gauge and hand plane.

I’ve been thinking of making a split handled hay rake, and this would be the perfect wood for making the tines of the rake.


Once square in cross section I use a small doweling jig and small palm plane to take off the corners. This is the same jig that I use to make arrow shafting, thus the long length.


With the dowel more or less worked to round it can be pounded through a dowel plate. I love this Lie-Nielson dowel plate, I’ve made my money back on it many times over any time I make pegs for securing mortise and tenon joints.


It does a good job as long as you keep the dowel straight as it is pounded through. I have a hole in my planing  beam for the express purpose of driving dowel through this plate.


Finally, the threads can be cut. I’ve tried soaking in oil overnight, and it does improve the thread cutting, but for the most part I don’t like to take so long to make a screw. That’s why its important to start with wood that handles thread cutting well in the first place. If there is a  bit of mangled thread that tears off, oh well, the screw usually still works just fine. I put a heavy coat of oil on the surface and go to cranking the thread box. They can be a bit tricky to get set up to cut the right diameter thread. I have a small mark on the threaded nut that indexes the cutter after sharpening. You don’t get a lot of thread cut before it needs sharpening, so it pays to be careful aligning the cutter.


The finished screws, compete with brass handles. I never get tired of the warm glow of brass rod, it really makes your work look high quality.


The hatagane in the back of the photo are the new ones, capable of clamping 24″ between the jaws. I also want to make a marking gauge that has a beam that can mark at 24″ and a long fence to ride the edge of the board, but it can wait until I get some of these panels glued up.

I keep on thinking I’ll have time at the end of the day to do this weeks joinery etude from project mayhem. I thought I would have time today, but making your own doweling by splitting from the tree does take extra time compared sawn lumber. I love when I can use lumber from the land though, it really gives that feeling of place to my work. Making these six clamps was a good day’s work. I squared my lumber for the practice joint and it’ll get cut tomorrow morning.

Anybody want to guess how much it would cost for six 2′ parallel  bar clamps by Jorgensen? How about ordering 2′ hatagane from Japan? I love my home made wood clamps.

Designing Fuigo (Japanese Box Bellows)

If you’ve ever searched for good examples of Fuigo, the traditional Japanese box bellows, you’ll quickly note the varying quality of construction skill. From the perspective of a cabinetmaker, screwing together a plywood box is functional, but so unsatisfying.

About the best thing you can find is a short video of John Burt:

I hadn’t really dug through the Daiku Dojo website enough to realize that Burt was part of the “Yataiki Nexus” about a decade ago. So, why build a fuigo with traditional joinery? Good question. If you didn’t have to worry about the time or the money, would your heart let you do otherwise? So much of why I want the bellows has nothing to do with making a quick cheap buck. Its all part of the experience in use, you know? And then compare the calmness of vertical grain paneling to the epic mess of ply veneer grain. I’ve been looking at cheap veneer laminate furniture my entire life, it never speaks to the soul.


Based on the general dimensioning that Danocon has used in his various fuigo I can tell that my fuigo will be a 3/4 scale, approximately. The vital statistics, about 36″ long x 21″ High x 12.625″ at its widest. Solid wood construction favours a box for the piston that is tall and skinny because it maximizes strength in the piston head board. All of the paneling with the exception of the piston head board needs to be as close to vertical grain as you can get. I don’t even think I would bother making 1/4″ paneling at 20″ wide if it wasn’t really solid vertical grain.

For a tuyere I’d like to use SCH 40 or 80 2″ black pipe, which measures about 2-3/8″ OD. The tuyere size and the size of the flapper valve cover determine the height of the “bird house” on the side that collects the air from the box and delivers it to the tuyere.


The size of the passive intake valves is some moderate percentage larger than the tuyere cross section.  For the side panels there is an interesting bit of joinery on the side with the bird house. The short side fits into a dado that runs the height of the long side panel, as well as a dado in the short side panel and a half lap where the two meet that locks the whole thing together. The side edges (the height) of the panels is slightly concave, such that the long side panels bow inward slightly when assembled. I’ll muse on how to determine a nice curve in a bit. These short side panels will be cut from a single piece of joined paneling 1/2″ thick x 20″ height x 24″ length. Why 20″ for the paneling height you ask…my sashigane is 20″ long. That’s how all of the dimensioning for the entire fuigo started, twenty inches on the sashigane.


This is the valve side panel, notice the notching in the lower corners. The notching along at the top corners allows the dados in both the top and bottom panels to be stopped grooves such that all of the side panel dado form a rectangle. These panels are thin, 1/4″.


Here are the bottom and top panels. The dado for the long side panels are bowed inward at the same radius of curvature as the long edges of the short side panels.  To form such a slight curve I’m going to make a ruler from oak, about 38″ long, and plane a curve of about 1/16″ deflection from the center. A plane whose sole is flat, with a blade projecting just slightly will actually cut a fair curve if you don’t take material off the extreme ends of the ruler. Start your plane about 1/2″ in from the edge and finish the cut short by the same amount. When the plane stops taking a cut you have a fairly even curve (once you remove the outer inch or so). If you want a curve with more deflection then set the plane for a slightly heavier cut.

John Burt describes how he cut the curved dado with a router, using the outside edge as a guide. With your curved ruler you could mark the curve on the outside edges of the top and bottom board. After planing in the curve your kebiki, marking gauge, will scribe the dado, yielding a dado whose sides are parallel and with an equal curve. I’ll be using a router plane to cut the depth and a marking knife to score the sides as I cut down. The whole point of making the ruler instead of just planing the curve into the side of the pal is that the ruler allows for the equal radius of curvature to be marked on the long edges of the short side panels.


The piston rod and head board have a double wedged tenon. This part must be prone to working loose over time and need tightening. Please note in the drawing that I didn’t reduce the dimension of the head board for the thickness of the bearing glass on the bottom of the box or for the gasket material. That’s dependent upon your gasket material and what kind of glass sheet you can find. I probably wouldn’t cut the glass until the side panel dado’s are cut so that you can cut the exact curves into the  side of the glass with your scoring tool running along the curved ruler you make.


Here are most of the pieces, not everything, with some dimensions. This doesn’t include stuff like the flapper valves or wedges, or the batons that go across the bottom board to give the fuigo “feet” and hold it off the ground.

This is a working set of plans, yet to be built, and I don’t show all of the pieces or the side view of the fuigo. So, if you want to build one you better watch the John Burt video about 100 times and do your own dimensioning. Still, I hope it is some help, we need to figure this stuff out.

Now I have a lot of work resawing and joining up for the paneling.

Making Sumitsubo (Ink Pot)


Making a sumitsubo reminds me a great deal of carving a wooden spoon. If you’ve ever used a chalk line and wondered whats with this Japanese ink line thingy, here’s the story.

So, I have these whale nokogiri that I’ve been practicing with, making some nice boards from the local lumber. Even though I like my chalk line, there’s some obvious advantages to having an ink line for laying out cuts on a log. The sumi ink is a beautifully clear deep black, and the thin silk line leaves a clear discriminating mark. Toshio Odate describes how these were often home made by a Japanese carpenter on a slow day or between jobs. Its worth the effort to make your own tools, you learn so much more about them, start with a deep understanding of the function of the tool.


I started by developing a pattern. For the top it only needs to have one side cut out, then flip it around a center line when marking and you maintain the bilateral symmetry. My pattern is designed 9-3/8″ long x 3″ x 2″.


With the pattern marked on the top and side of the blank I drilled four holes in the corners of what will be the mortise for the reel bobbin to maintain its location after the pattern is cut on the band saw.


I wasted as much wood as possible from the bowl with a forstner bit in my drill press. I’m using the power tools here to save time. If I didn’t have the bandsaw I would be making my rough cuts of the form with a carving hatchet.


I marked in my lines for the bobbin mortise with a piece of sand paper as flexible ruler and sawed the waste with a fret saw.


For the hand carving of the outside form and hollowing the bowl it really helps to have a couple gouges and a hook knife. I found the bent neck gouge absolutely indispensable for the deep hollow of the bowl on this sumitsubo.


This is actually attempt number two for the carving. The first had a bad split in the bowl and it broke (after I threw it against the wall). For this second attempt I used some different wood (Alder) and it still had a split! This time in the handle, where I let in a small butterfly splice to keep it from spreading any further. Carving a deep bowl like this is slow, and it was faster to repair the split than to start over again.


For making the bobbin I turned to my lathe and a 1/4″ pen mandrel.


I know from what Odate describes that most sumitsubo use a captured nut in the bobbin and a threaded crank shaft to secure the bobbin to the reel. I’ve fabricated my own square nuts and threaded shafting in the past for spinning wheel flyers, and decided that it was more work than necessary for this application. A small set screw, from threaded rod, is quite up to the task. Its not like you’re using this thing to reel in a marlin.


All that’s left is drilling the holes for the line. To get the angle I wanted on the hole from the bobbin into the bowl I used a piece of stout wire as a long drill.


But does it work? Works great! Though I think the silk line that I bought at my local yarn store was a bit fuzzier (and lower twist) than you’d want, but it still leaves a straight mark. God forbid I spend two dollars to order one from Hida Tool…I wanted hard stick sumi ink as well, but the art supply store in town only carried the liquid stuff.


Now that I have my sumitsubo I can saw up another log. Fun times! The chalk line was my favorite tool for a long time, but I think the ink line has taken its place. Now I just need to make one of those little bamboo brushes.

Next project: FUIGO

Making a Dai


I’ve been wanting a Naga-dai-kanna (jack plane) for a while now. For some reason it never occurred to me to use the 60mm blade that I already have for my smoothing plane and cut a new dai. So, that’s what I did. Before making my first dai I made an osaehiki nokogiri, the saw used to cut the blade bedding lines. Its very sweet, the first shavings from a new tool, after having to make a tool to make the tool. You get it, its a long and satisfying way to fill a shop with tools.

The plane measures 16″ in length, with a 40/60 distribution front to back from the cutting edge.


Let me say straight off, I think I made a lot of mistakes, hopefully you can learn from them. I didn’t have any plainsawn oak thick enough to make a dai from so I laminated one up from a couple of pieces of left-over red oak flooring. Its definitely not the same as Japanese white oak. There are these soft ring-porous areas in red oak (and the pieces in my dai) that I’m sure are not going to wear well in use. I made my layout with a marking knife when pen would have been more than adequate. The bedding angle is 38 degrees, and the koppa-gaeshi (small vertical line by the blade angle line) was made vertical. My primary source for this project was Toshio Odate’s “Japanese Woodworking Tools”.

Jay Van Arsdale’s making a dai was also very helpful, especially where they differ.


I started by mortising in from the top about three quarters of the way with a 1″ bench chisel, riding the bevel. I used the lines I had drawn on the sides as a rough guide and took care to cut shallower angles than was necessary.


Working from the bottom I cut between the blade angle line and the blade bedding line. I didn’t mark out the extra space for the shaving to pass through, figuring to pare it open once the blade was seated. I started by using the sokozarai-nomi to clean the waste, but in the end just started using a 6mm mortise chisel, bevel down. I cut to the depth of the koppa gaeshi.


Flipping it back over I pared back to the layout lines, but at a slightly shallower angle than necessary, and slowly pared down the angles. I started with paring the mouth opening towards the back of the plane (the surface with the chisel in the picture) so that if I pushed too hard and gouged the blade bed there would be enough material left to remove it.


The surfaces came out pleasingly accurate and smooth, ready for the side grooves to be cut.


I started the first cut and my little home made saw was binding badly. I remembered Jason had left a comment at some point about special files for hon-metate that raised a little burr on no set dozuki, so I made a light filing pass on the osaehiki-noko. It worked so much better with a little burr from filling. It was easy to make the cuts along the blade bed with that nicely pared surface to jig the saw plate.


For the blade angle line cuts I used the edge of the mouth on the sole to keep the blade vertical and on point and cut to my layout line on the top. I’m really pleased with how the new saw worked out, definitely worth making.


For cleaning the waste between the cuts I used a fine 1/8″ detail chisel, but an 1/8″ mortise chisel would have been better. This oak is tough stuff and a tiny delicate chisel can’t take heavy cuts without bending from the force.


From there I bedded the blade. What I did not notice was that the top of the blade is heavily concave. I didn’t leave myself enough wood when I marked the blade thickness line in layout. The process of bedding the blade was cut short because the blade became too loose before the top of the iron was fully seated to the block. I should have held the blade at least 1/4″ back from the mouth when marking the blade thickness line.


But what the heck, it took an okay shaving even with the poor bedding. I could feel chatter at the cutting edge, but decided to forge ahead and get the sub-blade holding pin drilled. Odate gives some obscure directions for figuring out the placement of the hole, but I decided to just trace the sub-blade on the blade angle line and use that line to place the hole. Perhaps I should have center drilled it first with a 1/16″ bit, my hole wondered terribly. At least the error is in a direction I can correct by filing.


Gotta love the warm glow of brass rod. Classy. Something else was bothering me though. I had been reading on Chris Hall’s blog “The Carpentry Way” about warped conditions on the back of new Japanese plane irons. Can you see how the blade is hitting the blade angle line with a gap in the middle? I didn’t do a good job flattening the back, and couldn’t of with so much concavity in the back of the blade without ruining the ura. I thought I had a basically decent plane iron, but cutting this dai forced me to really look at the blade in new light.


What ends up happening is that the blade wants to dig into the bedding on its back, effectively loosening the blade and changing the bedding angle. I ended up stoning the sharp edges off of the mimi where they meet the back, and it seemed to help somewhat with the digging in.  At this point I had learned a great deal, but didn’t have a plane that felt like it worked well. I had to sort out the poor blade bedding.


I cut the thinnest veneer I could with a handsaw from another piece of oak and glued it in the bed. You can see how after scraping in the blade once more I exposed the original wood on either side where the blade back was the highest. My god, what a difference a properly bedded blade makes, it even fixed a poor fit of the sub-blade against the sub-blade holding pin. The naga-dai-kanna does seem to take more time to condition the sole, but that is something I can now look forward to as I get to know the use of this new tool. And yes, it gets a board very, very flat.