Multi-Drawer Cherry Chest
June 1, 2014
A cabinet in naturally finished cherry is going to be striking, whatever
the design and size. It is just such a wonderful wood to work with and
rewards the woodworker with a second bonus: the color darkens and
picks up even more character as the piece ages, sort of like a fine vintage
These SketchUp images give me the ability to model the finished scaled piece in a 3d rotation--a fabulous tool
to give me some perspective to work with. I also took advantage of an
article (FWW# 141) by Becksvoort that demonstrates an easy, foolproof set of
calculations for doing graduated drawers accurately. There being so many
drawers in this piece, the size and spacing of them is critical to the
I made a pattern for the leg by
bending and clamping a thin batten on edge on the face of a 2x4 until I got
what I thought was an acceptable shape. I then traced the curve on the 2x4
and then cut it out with the band saw. From a photo of the pattern scanned
into SketchUp I was then able to incorporate the scaled leg pattern into my
I used double tenons for the blades to
gain some additional strength for these small parts. I also put a divider
stile into the back of the case in order to divide up the back panels into
smaller, more manageable pieces, with less overall seasonal movement.
By my last count, this chest was already up to 75 individual pieces (before glue-up),
excluding of course the d rawers and top. That is probably a new piece record
for me, even though I’ve built much larger items. Because of the 11 drawers
of different sizes and all of the runners, kickers and dividers, how to deal
with all of those pieces at glue-up and how seasonal wood-movement will
impact the finished case must be given careful consideration. These were
items that I had not given a lot of thought to when designing the piece, but
were now confronting me head-on when looking at the reality of gluing up
such a complex case. While SketchUp allows you to do some incredible
modeling in almost any perspective and zero-in on the look you want at
design time, at some point you will need to engineer all of those pieces
into a manageable glue-up. And, of course, SketchUp doesn’t care about
seasonal wood movement. These were lessons I learned on this piece, and will
certainly pay closer attention to these kinds of details with the design of
my next piece.
After cutting the dados in the legs
for the side and back panels, the curves were scribed onto them from the
pattern and then cut on the band saw. Most of the cleanup was then done with
a smoothing plane and cabinet scraper. All of the drawer runners and kickers
are mortised into the front rails/dividers and into the back center-stile.
While most of these joints will be glued at assembly, the connections with
the back panels fit tightly and will not be glued to allow for seasonal
movement within the oversized mortises. This will also somewhat simplify the
glue up process, while the case will still be stoutly held together by the
back center-stile glued joints.
I tried to look at each part of the case with the possibility of gluing up some of the items separately and then
adding them after the initial case assembly. Of course the 2 side panels and
the entire back panel could be glued up separately, applying the glue
sparingly only to the center of each panel tongue where they fit in the
upper and lower rails. This would allow the panel to move seasonally within
the grooves of both the center stiles and legs. I constructed the drawer
vertical dividers by attaching the front-end vertical piece with a biscuit
and then grooving the back center stile and ends of the dividers for a
biscuit. In this way, the entire divider assembly could be added later after
initial assembly. Although these dividers run horizontally within the
opening, I made then a little less than the exact opening height, so as to
allow them to expand seasonally.
Finally, the day of reckoning with the
glue-up process arrives. Will this go as smoothly as it can, what with all
of the thought and planning I tried to put into this moment, or will
disaster strike? With the help and steadying influence of my lovely wife
(“Help Me Wife, There Are Many Pieces!”), all went well. I glued up the
front panel with all of the drawer horizontal dividers separately the day
before and made sure it was square. I had also previously glued the side
panels into the back legs. All of the interior pieces (kickers and runners)
were numbered and a dry fit trial-run seemed to indicate that the best
assembly process would be to put them place with the case on its back and
then attach the front, starting at the bottom and working up. We each manned
a side on the case and mated each piece with its mortise(s), gradually
squeezing the front onto the sides and runners. Then we flipped up the case,
checked for square and added the clamps. It all went swimmingly.
Before starting the construction of the drawers for this chest, I remembered
an article in Fine Woodworking (FFW #224) on simplifying the steps to
well-fitted drawers. Because there are so many drawers (of several different
sizes) in this small chest, I thought I might look at expediting the
building process though learning some new techniques. All of the information
in this FFW article is helpful, but what was really a new idea to me was
fitting the drawer faces to the openings before actually constructing the
drawer. To do this, I had to build a new shooting board (something I’ve
considered doing for awhile) in order to accurately plane the drawer edges
(tops, bottoms, and ends). I was able to utilize by Lie-Nelson #164 plane
for most of this work, but having a plane designed just for shooting now has
crossed by radar for another possible acquisition.
The rear dovetails on all of these drawers will be simple
thru-dovetails, generally with only two pins being sufficient but adding an
extra as the drawer sides increase in width. Although I will cut off the
very bottom of the drawer back even with the top of the dado made for the
side to slide in, I leave the dados in the back so as to insure accurate
layout of the dovetails. I begin by marking the tailboard and cutting it out
using the 1 in 7 jig previously made for the band saw. Next the tailboard is
chiseled out and cleaned up in order to overlay it on the pin board for
accurate tracing of the pins.
The two boards are overlaid with the pin position accurately traced onto
the pin board and then marked down to the line for cutting with the handsaw.
I’ve tried several dovetail saws including a very nice Lie-Nelson, but I
seem to get the best results with this Japanese Shinwa Dozuki, which like
most of these Japanese saws, start cutting on the back stroke. This is
probably the hardest part of cutting dovetails for me, and this saw really
helps. Next, I clean up the ends of the pin board buy cutting it very close
to the line with the band saw fence.
Cutting a rabbit and marking the dovetails.
I read an article in Fine Wood Working (FFW# 219) on
simplifying construction of half-blind dovetails. Having labored through all
of the hand work before, I thought I give a try to some new techniques to
might make things a little easier and a little quicker. The first step was
to cut a shallow rabbit in the tailboard in order to make the alignment
easier and more accurate when tracing on to the pin board.
Next, using my trusty Lie-Nielson 1:7 dovetail marker, I mark the
tailboard and cut the tails using a 1:7 jig for the band saw. This gives me
nice straight cuts perfectly perpendicular to the saw table.
I chop the waste out of the tailboard with a sharp chisel and trim the
ends very close to the marking line using the bandsaw. Next I will
pare down the remaining small amount on the ends with a sharp chisel.
Before tracing the tailboard onto the pin board, I like to take a small
amount off the back of the board with the block plane. Then, when marking
the pin board, the knife will slightly under-mark the tailboard. At worst
this might cause a little more chisel work to get the perfect fit – but it
is always easier to remove more wood than to try to add some back. It seems
to help where my half-blinds are sometime most deficient: slight gaps
between at the ends of the tails. I started doing this about halfway through
the half blinds and the latter joints are definitely tighter.
Having the shallow rabbit across the tailboard definitely helps with the
alignment for tracing onto the pin board. I clamp the two together and make
my marks with a sharp xacto knife and then mark the waste carefully so as
know which side of the line to cut on (only need to do that once – that
sinking feeling, as Timothy Rousseau so aptly put it when a mistake is
The pin board is clamped into the simple jig I built from
instructions in the FFW article and checked with a square to be sure it is
perfectly flush with the top edge.
I use a trim router set to the depth of the tails to remove the waste near to the scribe lines.
Using my Lie-Neilson chisels, I finish cleaning up the pin boards. These chisels make it easy to get into the tight corners next to the pins.
The dovetails fit almost perfectly after cleanup with the chisels,
requiring little in the way of paring.
With having to build eleven drawers, I found this method of building
half blinds to be a real time saver. The piece was finished with
several coats of wipe-on tung oil finish, sanding between coats with up to 600 grit.
This cabinet won the Grand Champion Ribbon at the Northeast Washington State
Fair this past fall. Thanks to all the judges.