Making dowels




I made a few dowels this weekend for a project. A handy thing to know how to do, when you want some of a particular wood. A couple of tools in particular make this really quick.

First I cut square strips from my walnut on the table saw. Then I used a low-angle block plane to pare the strips down into octagon-shaped sticks. Stopping at this point might work just fine depending on the application. I could see the handmade rough octagon shape being a nice touch in some joinery.




Then I put one end of the stick into the drill press, supported the other end loosely with a hole in a piece of plywood, and used a shaver tool (like a Stanley Surform) to make it round.


It took me a little while to find the right tools to do this job, but now that I have the process down it will be quicker than going to the store and buying dowels. For very small amounts, anyway.

UPDATE: This is a crappy way to make dowels. some of them turned out to be out of round in a lumpy way that was unappealing.

Tools as a vector for expertise

Last year I bought an old wooden plane from a retiring woodworker at his shop sale. The man was curt. It was not a festive occasion, the mood in the shop pointed to his retiring probably more out of necessity than choice. A wide variety of tools was on offer, from borderline museum pieces to stuff probably bought at Harbor Freight within the year, but in all transactions the disposition of the man spoke clearly that you were lucky to have a chance at these tools, at any price.

So, without haggling I ended up with a few things: a level, a tenon marking gauge, a wooden-bodied plane. I mostly work with power tools, but there are plenty of situations, mostly involving big work pieces or weird angles, where a hand plane is useful. This one is a Stanley model 35, which means it could have been made any time between 1870 and 1940.

Screen Shot 2015-10-31 at 12.58.35 PM

Though it might look handlebar-moustachey, this isn’t a museum piece, or even an exceptionally good tool. It’s a decent tool, a money-maker, a thing in common use for decades, only recently turning into more of a fetish item for hobbyists who prefer muscle power to electricity on principle.

The sharpness of the blade makes or breaks the utility of a plane, and sharpening properly takes some practice. So I was greatly pleased when the blade of the old Stanley turned out to be sharpened recently and expertly. There are plenty of books, and even more web pages, on the right way to put edges on steel things. But even the best illustration lacks the descriptive power of the thing itself. From the angle of the metal, the portion of it that tapers down to an edge, the portion of that edge that is mirror-sharp, I can see what is important about sharpening and try to emulate what another craftsperson has done.

My IDE starts out with the same bits as yours, when we download them. But as we grow as programmers we customize our tools in ways that preserve our own solutions for getting things done. The preferences, scripts and template projects we carry around are records of how we dealt with complexity last time, that hopefully help make more things routine next time. What if our digital tools were designed to absorb and transmit our expertise in more and deeper ways? I want to see how great programmers sharpen their planes.


Wood storage

I got into the shop for a solid and blissful day this weekend to finish an organizing project. Spare wood I’ve been hoarding has been threatening to take over my small space for some time now. I finally got around to building a good solution.




  • 3 x 4″ by 4″x 8ft. ($10 each at Seattle Reuse)
  • 12 x 3/4″ diam. x 18″ precut black steel pipe ($5 each at Home Depot)
  • 6 x 3/8″ x 5″ lagbolts ($2 each)


Drill the holes for the pipes in each upright. The holes should match the diameter of the pipe and go about 3/4 of the way through each 4×4.

The tricky thing is that the outer diameter of these cheap pipes is not that consistent. I was hoping to use 3/4 holes and press-fit the pipes into the holes. Nuh-uh. The pipes were about 0.83 in diameter and were in no way going into 0.75 inch holes. I tried a 13/16″ (0.8125″) bit, and at this size the threaded ends of the pipes bit into the soft wood just enough that I could screw them in using a vise grips and get a tight fit. Now that I have a better idea of the tolerances involved, I could take a calipers to the store and come up with a solid plan for the pipe I’m buying. This time I lucked out.

Countersink holes for the two lagbolts in each upright, 1/2″ diam. by 1/4″ deep. Position the holes somewhere near the top and somewhere near the bottom.

Drill through-holes in each countersunk hole for the lagbolts, just under 3/8″ diam.

I used a bit that was 3/8″ minus 1/64″, or 23/64″.  The right hole size for the right amount of bite may depend on the softness of the wood. I drilled a couple of holes at exactly 3/8″ and though the bolts could not be pulled out of the holes, they could be rotated with a socket wrench when in all the way—not what you want.

Find the studs in the wall and bolt each upright to a stud. Then screw the pipes into the uprights, with a vise grips or pipe wrench, making sure to get each pipe at least two inches into the upright. That’s it.


I was using reclaimed lumber for this project. If I did it again I might look for high-quality, new uprights. The reason is that each pipe is acting like a knife when the shelf is loaded up with a weight. It’s a very dull knife, but nonetheless one that’s trying to split the upright in two. With good wood this should not be possible but I could see any pre-existing cracks around the hole getting wider until the upright breaks.

So structurally it’s a little on the funky side but practically I don’t think it will ever fail given the intended use. Once I had the pipes screwed in I lifted myself up on the low ones and bounced up and down a little, and they didn’t seem to give. Good enough for me.

I left the orange plastic caps on the end of each pipe to avoid scratching my wood supply on the threads.

Getting the holes straight is important. I started each hole on my small drill press, which only has about a 2 1/2″ throw. Then I made them deeper with a hand drill.

If I had it to do over I might go just 1/64″ larger on the pipe holes, then drill through the entire upright and screw each pipe all the way through until it reaches the other side. With the 13/16″ bit, the pipes were getting very difficult to screw in any farther after about 2″ of depth.

A small table for backpacking

Here’s a little project I did the day before a recent camping trip. Chaya was looking at a fold-up table in the REI catalog and thinking, wouldn’t that be great for hooking up morning coffee outside the tent. The REI table was $100 or something kind of crazy.

I remembered that Ballard Reuse, a great store a few blocks from the Madrona Labs shop, had some tongue and groove cedar boards in stock, and that’s what made the whole project click into being. T&G cedar is not something I have seen often—cedar is a great material but too soft for floors. I asked them about it at the store and they said it was probably from a ceiling or a sauna wall. Makes sense.

I got about 64 linear feet of the cedar, and put some lengths together to figure out how many boards wide my table would be. Then I spent a while picking out long-enough sections with no nail holes or cracks. These formed the bulk of the parts:

  • tongue + groove boards, 18″ long and enough to make 20″ width
  • 2 hardwood sticks, 1/2″ by 1″ by 18″
  • 4 hardwood dowels, 1/2″ by 12″
  • 4 steel T-nuts, 1/4″-20 tpi.
  • 4 hanger bolts, 1/4″-20 tpi. by  1 1/2″


Finding the right hardware can make a project go together very easily, when you’re just trying to make a thing that works and not doing “fine woodworking.” In this case the t-nuts and hanger bolts provide all the rigidity this small table needs, and allow for easy assembly.

The only tricky step in making this is boring out the dowels to insert the hanger bolts. Doing it with a hand drill, it would be hard to get them in anything like straight. I used my drill press to drill holes 15/64″ diam. by 3/4″ deep, then added some epoxy to the wood screw ends of the bolts before screwing them in.

Then I drilled the holes in the ends of the sticks, 17″ apart. Marked centers off the same distance on the underside of the cedar boards with an awl, and put in the T-nuts. Quickly marked the corners A, B, C, D for assembly. It should go together either way, but when you’re setting up camp it’s nice just to follow instructions.


After this is done you can screw the legs into the end boards through the stretcher sticks,


turn it over, and enjoy a cold beverage.


Process: drawn arc to contoured board


Over the past couple of weeks I took some snapshots as I worked out a process in the woodshop, a problem I was puttering on a few hours at a time: how to turn an arc drawn on my workbench into a long board, contoured with that arc. The solution seems laughably complicated. But the result is very accurate and looking back over it, I can’t think of an easier way to get it done.


Draw an arc.


Make a solid section of the cylinder defined by the arc.


Make a solid section of the negative space excluded by the cylinder.


Make a rigid sled with sides to constrain a router, and with the negative section at its short ends. Place the sled on a base with the positive section at its short ends.


Make rails to keep the movement of the sled aligned along the arc.


Attach a long board to the base.


Mount a router in the sled. Mill away the entire top surface of the board.


This process will generate a curved board and many wood shavings.


The contour of the board’s curve will match the arc drawn at the beginning of the process.




An unexpectedly nice board

Yesterday I went into the shop to work on an fun project, a shelf with some kind of cubbies for my standup desk. I thought I might use this big old board I found in the yard when we moved into our apartment. It was nice and straight, but covered with years of mold and gunk. About two inches thick, it looked like it could be resawn down into a lot of good one inch thick stock.


So I got out the hand plane and cleaned it up a little just to see what was underneath.


What was underneath turned out to be really special. I’m not sure whether it’s the red or white species but this is definitely oak, about 80″ by 10″ by 2″ of it. Maybe a little too nice for my standup desk. What I’ll probably do is make a plywood version—then if it fits well and turns into a piece of furniture I want to have for many years, I’ll do the oak one.


Woodworking is interesting this way—the materials can surprise you by revealing hidden properties that elevate or diminish them. You can run into a nail or a knot that can ruin your day.

Programming is more predictable. Bits can ruin your day, but not by having unexpected properties. In general, this is obviously a good thing, but I think that in my case it sometimes perpetuates the illusion that I’m in control of things, where woodworking keeps me more grounded.