A full discussion of dovetailing is beyond the scope of this series, but here are several tips pertaining to drawer making.

Much of the success of dovetailing depends on accurately transferring the outline of the tails to the pin board (or vice-versa, if that is your preference). The bottom edges served as references when fitting the front and shooting the sides square, and continue that role as a heavy, flat board, squared to the front piece, aligns the sides to the front while marking for the pins.

Angles for short dovetails should be steeper than usual to improve their appearance and possibly their strength. The tails for these drawers are approximately 5:1 slopes. My usual ratio for medium to larger joints in hardwoods is 7:1.

Layout the joints so the groove for the bottom will be fully within the lowest tail at the front. At the back end, there is a large tail at the bottom which also fully contains the groove, while a tail at the top allows for a neat chamfer.

The grooves to fit the drawer bottom are cut after dry fitting the front and sides and ensuring that the reference edges (the bottom edges) are aligned. For these drawers, I made the grooves 1/4″ wide, barely 3/16″ deep, and 3/8″ away from the edges. Keep the grooves sufficiently shallow and away from the edges to avoid weakening the sides, though not so shallow as to allow the bottom to slip out if it distorts a bit.

Note that the bottom edge of the back piece does not end with a customary squared pin. The little shoulder below the lowest pin allows the bottom of that pin to have an angled surface which has contact through the full depth of the side piece. A squared pin lacks the added strength of the angle, and contacts the side piece only in the thickness remaining lateral to the groove depth.

I prepare the back piece slightly wide at its bottom, cut the joints, loosely dry-assemble the joints, then mark the back piece so I can plane it, after disassembly, to align with the top of the grooves. This method separates the process of dovetailing from aligning the bottom edge of the back with the groove, affording more control in workmanship, as well as making stronger joints.

Next: adding the false front and fitting the drawer. Stay with me, we’re getting to the cool part.

For the sides and back, it is best to use straight-grained quartersawn stock for the sake of dimensional stability. This wood will usually have a plain appearance and offer a contrast to the drawer front. Light-colored species are typical, such as hard maple or yellow poplar.

The thickness of the sides varies according to the size of the drawer and the load it is expected to carry. I prefer sides to be a bit chunkier than those favored by many craftsmen. For the small drawers in this project, I made the sides 7/16 inch thick. The backs are a fat 9/16 inch, not much less than the front, to allow for stronger joints at the back corners and to create a more balanced drawer as it is withdrawn.

It is helpful to select parts so the outer surfaces of the sides can be planed front to back after assembly. The height of each side is ripped to that of the drawer front where it will be joined. I set the table saw fence using the front as a gauge. The length of the drawer should allow for safe clearance at the back of the case, considering any anticipated shrinkage of the case during the dry season, as well as for a small projection of the bottom past the back of the drawer. The length of the back piece exactly equals that of the precisely-made front, maybe plus a hair, but never shorter.

Triangle marks are invaluable to keep parts organized. A number inside the triangle section is helpful when constructing multiple drawers.

Because I used Port Orford cedar, a rather soft wood, in this project, I was concerned that the bearing surfaces would wear over time. If the sides of a drawer are too thin or too soft, drawer slips, a traditional solution, effectively widen the bearing surface as well as prevent the groove for the bottom from weakening a thin side.

In this project, I wanted to try a different, perhaps cleaner-looking solution, so I glued 3/16 thick strips of hard maple to the bottom edges of the drawer sides and planed them flush. In retrospect, I can’t say this was any easier than making drawer slips, but it worked out well.

Join the sides to the front and back using whatever dovetailing method you prefer. The bottom edges are the references. Here’s a key point: set your cutting gauge to produce a pin depth that is slightly less than the thickness of the sides. Don’t go crazy measuring these tiny amounts; they’re not critical. Call it less than a 32^nd. Later, after assembly, the sides therefore will be proud of the front and back. They will be planed flush to the end grain of the pins as you reap your reward for the care you invested in fitting the front to the case. Beyond that, only minimal, judicious planing, if any, may be required to achieve the sweet fit that you seek.

Next: several pointers on dovetailing as it pertains to drawers, and a different method for dovetailing the back to the sides.

The first, most critical step in making this type of drawer is sizing the front. The drawers in this project are about 14 inches wide, 3-4 inches high, and 12 inches deep. For the fronts, stock is face jointed and thicknessed to about ½ inch. “False fronts” will be applied later to bring the final thickness to about 11/16 inch. (The methods described here are applicable whether false fronts or customary half-blind dovetails are used.) Each piece is ripped to a hair less than the height of its housing and crosscut to slightly larger, about 1/32 inch, than the housing width.

Unplug the machinery. The following hand planing is best done with a shooting board. [Here is my shooting board and how I use it for end grain on any size board and for long grain edges on small boards.]

First, the bottom reference edge of the drawer front is planed straight and square and identified with a mark. Then the left end is shot to match the left side of the opening. Progress is tested by offering up the piece to the housing, resting it on its bottom, and checking the left edge. (Photo, above.)

Now it’s time to get very careful. This step is probably the most important point in your drawer’s success. Bit by bit, the right end of the drawer front is shot so it just barely makes it into the width of the housing. It should be snug! 

Finally, the top edge is planed so it comfortably fits in the height of the opening and is parallel to the upper edge of the opening. By keeping this gap small, 1/32-1/64 inch, it is easy to detect any deviation from parallel by eye. Note that the top and bottom edges of the drawer front are not necessarily exactly parallel to each other at this point because the piece has been fit directly to its opening.

This is incremental work. It is nice if the drawer opening is a perfect rectangle but this is not assumed. Any deviations are accounted for by fitting each drawer front to match its opening. With this method, errors do not compound as the project progresses. The drawer front prepared in this way is now the reference for constructing the entire drawer.

I do not make the height a snug fit as is sometimes taught. I believe that is a waste of effort since the finished solid wood drawer must not fit tight in height because it will bind when the humidity rises. The “air tight” drawer, in my opinion, makes no sense; it won’t work. I also do not shoot the front to produce tapered edges for fitting in its opening, because I want to work with square reference edges when I cut the joinery.

Once you have thoughtfully prepared the case and fit the drawer front accurately, there is not much left from here on that is at high risk of going wrong, as long as you are patient.

Next: making and joining the sides and back.

Thoughtful planning and careful workmanship in constructing the case that will house the drawers will pay off later. The web frame construction in this solid wood design, only one of many ways to create a drawer case, will serve to illustrate some key principles. In this project, each runner is set in a dado in the vertical partition which forms the side of the drawer housing. A tenon at the front of each runner is glued into a mortise in the front divider. The runner is screwed to the partition near the front and is slot-screwed near its back end. The rear end of the runner has a tenon which is fitted without glue into a mortise in the back divider. This arrangement allows the case sides to move unrestrictedly with seasonal moisture content changes.

Whatever the form of case construction there are some important points to monitor. The width of the drawer housing should slightly widen toward the back of the case. This allows the drawer to be pushed in and pulled out smoothly without binding. Ideally, as the drawer is pulled out almost to its limit, the sides will gently tighten against the housing. It is futile to measure this tiny widening with a tape or rule. Instead, cut a piece of scrap or use a pinch rod setting so it just fits the width (or gently binds) at the front of the housing. Then slide it toward the back where you want it to “release” and slide freely. The clearance is perhaps 1/64 inch; I don’t measure it. At least ensure that the housing does not narrow toward the back.

In the photos below, getting a bit ahead of ourselves here, I’m showing how a fitted drawer front laid flat is snug against the sides at the front but has clearance at the rear of the case. (Fitting the front is covered later in the series.)

It is easier than it might seem to achieve this sort of tolerance. For this project, I cut the joints, then dry fitted the vertical partitions and performed the testing process as described above. Then I disassembled the case and simply hand planed away some thickness on the inside surfaces of the partitions according to the indications of the testing. I reassembled the case and retested. After the case refinement was completed, I ran the dadoes and constructed the web frames.

A few more things require attention. The surface on which the drawer rides should be free of twist. This is mainly controlled by cutting the dadoes for the runners symmetrically in each partition. The front to back consistency of the height of the housing is less important but the height should not decrease toward the back. The front opening ideally should have four 90 degree corners, but don’t worry, small errors can be compensated when sizing the drawer front.

These same general principles can be applied to fitting solid wood drawers into other case designs, such as frame and panel, plywood, and veneered constructions, though the planning steps required to implement them will be different. For this series, I will stay focused on one example of a solid wood project.

Next: how fitting the drawer front is the key step.

There are so many ways to make drawers that a book would be the right medium for a comprehensive discussion of the topic. This introduction begins a series of posts, not likely to be contiguous, which will focus on one method for high quality, fairly small size drawers suited for a craftsman’s best projects.

Sometimes I wonder why we woodworkers bother with the niceties of fine drawers. I still warmly remember the hectic weeks with a newborn baby in a new house when I stole any minutes I could to build a large tool cabinet for my new shop. Six drawers, nothing too pretty: rabbeted plywood, bottoms running side-hung in dadoes, glue, screws, and feeling tired but happy. Now more than two decades, college expenses, and a lot of woodworking later, they still run smoothly. It would be nice if everything worked this well.

Nonetheless, at the other end of the aesthetic spectrum it is certainly possible to combine function with beauty. A logical process, with special attention paid to the critical junctures, will produce enduring, exquisite drawers. This series is based on traditional methods, but I will feature some modifications that I use because they make sense.

This is not the only way to make fancy drawers, nor do I propose it as the “best” way because that judgement depends on function and aesthetics which are ultimately the provinces of each craftsman for each project. For making high-end drawers, as with almost all of my woodworking, I employ machines and hand tools, though the latter predominate and certainly are used for the precision steps.

The next post in the series will address the fine points of case construction with regard to drawer fitting.

Trick question. It depends. This post will discuss factors in the amount of camber in the edge of a plane iron with attention to an under-appreciated trigonometric quirk.

The slight convexity or “camber” in the edge of a smoothing plane iron should allow the production of airy shavings that are thickest in the middle, say .001″, and feather out to nothing at a little less than the width of the blade. This produces imperceptible scallops on the wood surface and avoids square-edged tracks or “gutters“.

A similarly small, or perhaps a bit more, camber in the edge of a jointer plane blade allows one to bring down the “high” side of an out-of-square edge without tilting and destabilizing the heavy plane. The camber should be positioned at the center of the blade projection so the plane can be shifted toward the high side of the board’s edge to remove a slightly thicker shaving there.

For jack planes, more camber lets this workhorse take thicker shavings without producing gutters. The more pronounced camber also makes it easier to direct the plane’s cut at the high spots on the surface of a board being dimensioned.

When grinding and honing a plane blade, I check the camber by setting the blade’s edge on a small aluminum straight edge and holding it up to the light to look for the tiny gaps that gradually enlarge from the center to the sides of the blade. (I never measure this amount so I cannot answer the question posed in the title of this post.)

Now some trig. Let’s say the camber – the depth of the convexity of the edge – is .005″. When this blade is installed on a 45 degree frog in a bevel-down plane, the actual functional convexity is reduced. Think of it this way: if the blade were laid flat and you viewed it toward the edge, there would appear to be no camber at all. The functional camber is reduced by the sine of the bed angle.

sin 45* x .005″ = .0035″

Look what happens in a bevel-up plane with a 12 degree bed:

sin 12* x .005″ = .001”

Therefore, I sharpen more camber into a blade for a low angle bevel-up plane than for a bevel-down plane to achieve the same functional amount of camber. The camber that you observe sighting 90 degrees to the face of the blade will mostly disappear when you install the blade in a 12 degree-bed, bevel-up plane and sight down the sole to observe the camber. Compensate for this by being generous with camber in the sharpening process. A more direct approach during the sharpening process is to check the camber against a straightedge with the blade tilted at the bed angle.

Again, I do not measure these things but rely on my eye, experience, and especially feedback from the work. Of course, sometimes I’m off, usually by over-cambering. However, since the middle of the blade is thus destined to dull first, it is easy to reduce the camber on the next honing.

There are undoubtedly other factors affecting shaving thickness, such as blade sharpness, blade edge deflection, and wood grain, so it is most important to monitor the performance of the plane and make adjustments when you resharpen.

You can use trial and error or a set of leaf gauges to work this out to your liking. Like just about everything else, there’s more than one good way. Use the principles and find your way.

The previous post discussed preparing the edge joint for thin boards.

The second problem involves clamping the joint. Because these are thin boards it is desirable to align the edges as accurately as possible to avoid a significant loss of thickness in the final product. More vexing is the tendency of the joint to explode when clamping pressure is applied. This problem is akin to trying to control the writing tip of an 8 inch long pencil held only by the eraser.

I like #20 biscuits for aligning 3/4″ edge glue ups and #0 biscuits could work for ½” boards, but for the 11/32″ boards shown here I would be concerned about biscuit swelling and show-through. Splines are a hassle which I gave up on long ago. The same goes for those elaborate clamps that use wooden battens to apply pressure on the face of the glue-up while the clamp screws apply pressure to the joint. One option that I have not tried but looks good is the Plano Clamping System.

Here is a simple solution that works for me. The boards rest on 3/4″ MDF or plywood platforms. After applying glue and bringing the joint together, I apply just light pressure with the bar clamps. Then I clamp small blocks, with cutouts to vault the glue squeeze-out, across the joint near the ends. I push or tap near the middle of the joint to align it there. Then I apply final pressure with the bar clamps.

There are surely many other methods for edge joining thin stock, notably from instrument makers. I’ve described simple, shop-tested methods that I use. Best wishes for your woodworking.

Edge joining thin boards, in the 1/4″ to ½” range, presents two special problems, both easily surmounted with the methods described here. These are usually fairly short pieces of wood, such as for panels and drawer bottoms, which permit alternative methods.

First, it is difficult to plane a straight, square edge using the usual procedure of clamping the board in the front vise and running a bench plane along the edge. The narrow edge provides little purchase to balance the plane consistently square to the face. The solution is the shooting board. I bring the two boards together, like closing a book, and align the working edges. Then I set the pair on the shooting board platform with the edges extended slightly beyond the shooting board’s running edge. The plane is, of course, used on its side, but the sole only touches the edges of the work pieces. Hold the boards firmly or use a clamp. Planing the two edges simultaneously in this manner negates any slight discrepancy from square.

In setting up for an edge joint, care must be taken to match and orient the boards properly. It may not be possible to meet all of these criteria with the available stock, but the first two should not be compromised.

• Join edges with similar cross-sectional grain orientation, rift to rift, quartered to quartered, flatsawn to flatsawn. Dissimilar edges, such as quartered and flat, will seasonally move differently in thickness to create a step at the joint surface and possibly stress the joint.
• Avoid figure runout or dissimilar figures at the edges of the boards. For example, do not juxtapose cathedral figure lines running off the edge with riftsawn straight figure. Make it look good. This is partly related to the above.
• The surface grain of the boards should run in the same direction to facilitate planing the glued-up board.
• The grain on the edges should run in the same direction when the boards are “folded”. This will make the edges easier to plane simultaneously, but is not a factor for thicker boards that are planed separately.

For the enduring question to plane the edge straight or concave, my simple answer is this: I aim for a straight edge allowing the least possible concavity but zero convexity (a one-sided tolerance). The ultimate test is to stand one board on the other and swing the top board. It should barely pivot at the ends, never in the middle. It must also not rock due to twist in the edges. Finally, a straight edge placed along the surfaces should predict a flat glue up. (Contrary to the appearance in the photo at right, I have five fingers on my left hand.)

When joining flatsawn boards I usually look for the nicest appearance and do not worry specifically about whether the heart and bark faces alternate or not.

By the way, it took me longer to write this than to make an edge joint. In the next post, I’ll describe a method to solve problems in clamping the joint.