What I want to achieve with this design is a desk that can be broken down and reassembled with few tools. This leaves me with just a few choices when it comes to attaching the legs to the top. Metal fasteners tend to fail with long term use and require the right size wrench or key. So I wanted to come up with something simpler that would stand the test of time. Attaching the legs to a tapered dovetail batten would allow the desk to be broken down into three, relatively flat, assemblies with nothing more than a hammer (or some other hard, dense object). But how much taper would be the right amount to lock the battens in place without having pound them into submission.
Time to cut another test joint.
My first instinct was to taper the battens by 1/4" over two feet. This would have worked, but it took a lot of blows from the lump hammer to seat batten. In the end I went with twice as much taper, or 1/4" per foot. But I'm getting ahead of myself. Here's the process I followed to cut the joint.
I started by marking the depth of the dado on both the batten and the underside of the sample desk top. Then I lined the center of the batten up on the center line of the dado and clamped it into place. The wide end of the batten was sticking out just slightly from the back edge of the desk top. I used a square to transfer the widest points of the batten to the knife line at the bottom of the dado. Then I set a bevel to match the angle of the batten and used a knife to scribe the angled walls of the dado.
I used my track saw, set to the same angle I used to cut the batten and a shallower depth, to cut the walls first. I followed up by adding several extra kerfs to make the waste easier to remove. You may notice in the photo above that I missed my knife line on the left side of the dado. This ended up costing me a lot of time when adjusting the fit. I should have adjusted the position of my track and made a second cut rather than trying to fix it by hand.
I removed ("bashed out" might be a better description) the waste with a chisel, turned bevel down, and a mallet. The surface left behind is a little rough, but my router plane cleaned it up and brought it down to the final depth I scribed with the marking gauge.
The first test fit went well, but still needed some adjustment to get the test batten fully seated in the dado. I'll discuss that process in the next post.