Tuesday, November 5, 2013

Hand-cut Dovetails on Compound-Angles

Before dovetailing the drawers for Angle Madness, I wanted to work through a procedure to easily "calculate" the compound angles I'd need for the drawers.  I say "calculate" in quotes because I wanted a procedure based exclusively on the use of bevel gauges reading angles in-situ.  This will be useful if you don't have your slide-rule handy...

For the drawers, I need two types of angles: the rear has inclined tail boards with a vertical pin board. The front of the drawers has both inclined tail and pin boards.  Ultimately, they are the same thing since a "vertical" board still has an angle, but there are some shortcuts we can take when only one board is inclined.  Things get only marginally more complicated for a joint with both boards inclined.  Seriously! Only marginally more complicated.

Even if you never plan on doing compound-angle dovetails, the portion of the procedure used to determine the projection of the boards on each other would be useful to make a simple butt joint between angled parts.  But don't stop at the butt; the dovetail is easy!

This episode repeats the procedure 4 times for the different joints. Hopefully that reinforces the procedure without being a snoozer.

Two years ago (already?!) I did a short series of videos on hand-cut dovetails. I'm the first to admit I'm not a great dovetailer, but the methods are there for the ho-hum pins-first/tails-first decision, but also for houndstooth dovetails, mitered dovetails, and others.  This blog entry is the first for the series; the table of contents at the top will show you the others.  Uhg, it was 104ºF in the shop during those videos. There's a reason for wearing all black!!

The next episode of Angle Madness is also partly recorded so I hope to not disappear for a month again :)


  • ChrisHasFlair said...


    Good video on the basic joints. How about one on cutting joints in two boards with curves in three planes (a la Insanity 2)


  • Shannon said...

    Well done Paul-Marcel. I scoured through all kinds of resources to find a hand method and every time it started with boards already cut to the appropriate compound butt joint. I finally found a solution in a timber framing manual from 1876 that really describes the process you just went through. Its one of those stupid simple things that we have forgotten with all our fancy angle charts. This was really well explained. The taped joints was brilliant!

  • HalfInchShy said...

    Thanks, Shannon! Glad the procedure matches something done in the past. I never looked around to see how since I wanted to figure it out myself; helps with retention :)

    Someone gave me an angle chart book once long ago when I started Angle Madness. When I wrote the program to calculate the angles, I tried to use the book to verify the calculations. Could not wrap my head around that book of tables!

    When I was doing one of the drawers for Angle Madness last night, I was using the blue tape on the end grain. Soooo much easier! :)

  • Brian said...

    I missed the video part where you were dove-tailing with the brace, did you accidentally cut out that part of the video?

  • HalfInchShy said...

    I did cut that out, Brian. I only had the camera on 60 fps. Since I was jacked up on 4 shots of espresso to get the speed I needed, I actually needed 240 fps to truly capture the moment. It was just a blur, not unlike the Tasmanian devil going through a taco stand after a hard day.

  • Rob Horton said...

    Quite possibly one of the finest instructional videos I've seen in a long time. Teaches not only "how", but also the "why" and "what happens if you don't..."

    My only refinement would be in the pencil technique used at the very end. If you've got a number of these to do, why not make a "magic block" a la Tommy Mac's serving tray?

  • HalfInchShy said...

    Thanks, Rob!

    I assume you mean the technique for stealing the angle to draw the horizontal lines. It turns out the angle is sitting in your hand before you draw it out and you had it on the third "transfer" bevel earlier in the procedure :)

    This video was the third recording I did. The first one, uhg, it was obviously not a day to be in front of the camera :) I'll leave it at that :)

    The second recording was good, but I explained the transfer bevel short cut you can use in place of drawing the horizontal; while editing it, it seemed clear it would completely lose someone on why that it correct or why it isn't one of the existing bevels. So I opted to re-record to both add more repetition (yeah, it's repetitious by design :) and to steal that angle in a way that makes it obvious what you are stealing (then prove it doesn't match the reference bevels).

    Someone wrote me with a lot of questions; we had a back and forth then he went to play in Sketch Up to prove things to himself more. I recorded a private video as a video-mail for him but decided not to send it and instead record it as a mini addendum. It will show you where you can easily steal the angle without a pencil mark and I came up with a visual demo that helps solidify the idea that the miters can't match the existing reference bevels if those are inclined. (So, thanks, Etienne!)

    I type too much :)

  • neilc said...

    excellent tutorial, Paul-Marcel.

    One of your best!


  • HalfInchShy said...

    Thanks, Neil! One of your best comments! :) haha

  • Ed said...

    Should you have used the 3rd bevel gauge at 22:15 instead of the tail board bevel gauge?


  • HalfInchShy said...

    Hi, Ed,

    It's true that if you hold the reference gauge for the tail board's incline right up flat against the cut miter that there's an error. It amounts to the width of a pencil line (in fact, Etienne sent me email about this very thing; his questions were the reason for the follow-up tidbits video).

    If you want the error to go away, it is very easy to do with that reference gauge (the one used to incline the board, not a third transfer bevel).

    Open up this picture in a second window; it is a still-frame marked-up from the video to make the next paragraph easier to read :)

    When tracing the horizontal line on the end grain, don't hold the gauge so the body of the gauge is parallel to the cut miter. Instead, orient the body of the gauge so it is perpendicular to the top edge of the board. In the photo, the magenta markup shows how the center of the reference gauge is not parallel to the miter cut but rather perpendicular to the top of the board. You then use the inside edge of the gauge's blade to mark the line (here shown in green).

    If you rotate the gauge to be parallel to the miter cut, you'll hardly notice a difference. When I cut my drawers, I marked the way this photo shows. On the video, I was drawing lines with a fat Sharpie for clarity so the line was an order of magnitude larger than the error :)

    Try marking it both ways and you'll barely see a difference. Much easier than setting up a third transfer gauge.

    Mathematically, you are correct, though.

  • Ed said...

    Thanks for the response.

    I did notice the way you held the gauge when I watched the vid yesterday. I couldn't come up with a well worded way to describe keeping the gauge in a plane that is vertical and perpendicular to the face of the tail board and then projecting the horizontal lines onto the endgrain from a viewpoint perpendicular to that plane. See? Still can't.

    Your response is more concise and the inner edge of the blade is very close over 1/2" thick side.

    Thanks for all great info in your vids and blog. This site has been very helpful over the last few years. The MFT vids(I waited 10 days for my Qwas dogs before I set up my MFT), TS75 splinterguard, early Domino and later calibration vid, OF1400 vids, Knew Concepts fret saw and I could go on. All have resulted in purchases and/or mods that improved the use of the the tools.

    You are the first site I come to for any review. The latest Valfor Tools vid is a case in point where you provide a well thought out calibration method that makes the backlash a non-issue.

    Thanks Ed

  • HalfInchShy said...

    Thanks for the great compliments, Ed! Appreciate it!

    Completely agree with trying to word these projections; I'm not a math major!