Thread: Dovetail practice - WIP

Keep those words and photos coming Javali, looks just like the type of woodwork I love and practice/practise.

Good stuff Javali

The Timber isn't Tassie Oak, most likely it is Blackbutt or Box Brush amongst others.

Anyhow top little project and keep us informed.

Like the panel clamps

Studley

Pic 1 - The boards are too wide for my square. I was a bit skeptic about using a ruler with the square, but it seems to have worked.

Pic 2 - The boards are also too wide for my shooting board. To prevent tear out I clamped a small backing block to the board.

Pic 3 - It does come up square. Some light is visible, but the camera seems to pick up more of it. It is not that bad in real life.

Pic 4 - I used my new Ryoba saw for trimming the other end down. It is not as easy as I thought it would be. The cut comes up as smooth as a baby's bum, and about as straight. I guess I am a western-saw type of person.

Pic 5 shows the dimensioned board. Only 3 more to go...

Pic 1 - Sawing the tails. The cuts have to be square and straight, but as the tails will be used to mark the pins, the cuts do not have to be too accurate. I am very good at making inaccurate cuts.
Sawing two boards together is time effective and helps keeping the saw square.

I chop (pic 2) and pare (pic 3) the waste.

Pic 4 shows the tail side.

The tails are used for marking the pins. (Pic 5)

Some of my Studley timber looks remarkably like that, with pencil marked SG. Might that be spotted gum?

Tails look good Javali. Keep up the pics, and thanks.

Tex

After copying the marks the lines are extended to the bottom of the pins (pic 1), and the pins are cut (pic2). Unfortunately this cut has to be accurate or the pins do not fit the tails.

The bulk of the waste is chopped. To avoid massive tearout I make a cut, about 3 mm deep, along one side (pic 3), leaving enough material to support the waste when chopping from the other side (pic 4).

Pic 5 shows the four dovetailed boards.

The sides are dadoed to fit matching tongues in the runners. The dado ends are hidden behind the dovetails that hold the dividers.

Pic 1 shows the setup for cutting the dados. I use an auxiliary fence nailed to the board for cutting the dados. The cut is not on a visible surface, so no need to worry about covering the nail hole later.
To reduce tear out, I chiseled out the end of the dado, and scored its sides. I do not trust the spurs on my #45 to do that correctly. I also planed towards the rear of the board so that any tearout will be invisible.

Pic 2 shows the plane.

Measurement introduces errors (at least when I do it), so instead of measuring again on the other side I transferred the location of the dados from one board to the other (pic 3).

I cut the tongue using the same #45. I learned the idea of using the dowel to help keep the plane upright from Alf. (pic 4)

The fit is too tight - probably because planing across the grain leaves the dado a bit too narrow, so I planed the tongue down a bit. (pic 5)

Great thread Javali. Keep those pictures coming.

Yeah keep 'em coming Javali...

HH.

I forgot to mention that I am aware that the tongue is not centred on the runners. It means that I do have to make sure that all the runners are aligned the same way, and always remember what side is my reference face.

So pic 1 shows how I use a mortice gauge to transfer the dimensions of the tongue to the tenons, and more important, to the mortices. Some of you may notice that the reference side for copying the dimension is not the same used for planing the tongue. This is not a clever trick. I almost stuffed all the runners with that. Luckily I do measure twice before cutting and noticed the mistake. I forgot to take a photo of the second try.

Pic 2 shows the runners. The runners (excluding the tenons) are slightly shorter than the distance between the dividers and the back rails, to allow for movement of the box sides. For the same reason, the tenons are only glued to the dividers.

Pic 3 - The length of the dividers is copied from the top board. this ensures that any inaccuracy in measuring the boards (they are about 1 mm shorter than the design) is transferred to one the dividers.

This is extended to the second divider and to the back rails (pic 4), and they are trimmed to length (pic 5).

Traditionally, the dividers are attached to the carcase using a sliding dovetail. I do not know if the traditional sliding dovetail spans the whole width of the divider. Trusting my abilities as much as I do, I decided to make them only about 10mm deep.

So, the small step in pic 1 will become the sliding dovetail.

Dimensions for the mortice are copied directly from the tenon (pic 2), and the mortice is marked (pic 3).

Sawing the sides of the mortice (pic 4) helps chopping the waste out (pic 5).

The beutiful thing in the last photo is my 1/4" firmer/mortice chisel. It is as nice to use as it is to look at. I think that the previous owner used it as a dog toy before putting it for sale. I guess that with a little work it can be quite a good chisel. All it takes is replacing the handle and getting a new blade. Anyway, I have a birthday coming soon, and would not mind getting a nice 1/4" chisel.

The only reason for marking the dovetails on the dividers is to give a general idea as to how deep the saw cut should be. (pics 1 and 2)

After they served this purpose they are ignored. (pic 3). I removed material until I got nice and crisp edges, as in pic 4.

Having been lax about accuracy comes with a cost. with the dados marked near the planned position and the tongue not being centred on the runners, I cannot rely on the plan for the exact length of the vrtical dividers.

To measure from the parts I already had, I marked the edges of the runners on one of the sides of the box (pic 1), and transferred the measurements to the dividers. The result is a perfect match. (pic 2).

Pic 3 shows the set up for copying the shape of the dovetail to the divider. Pic 4 is the marked socket after sawing the edges. The scocket is then chiselled out. (pic 5).

Gluing the thing up proved to be much easier than I expected. This is the current state of the project. Future posts will come at a much lower rate.