Home

Furniture Gallery

Individually Styled and Crafted

Fine Hardwood Furniture

by

Andrew Pitts ~ FurnitureMaker

Work in Progress
Television/Snuff Bottle/Jewelry Cabinet

Article completed 14 July 2008

Disclaimer: This discussion chronicles the making of a piece of furniture in my workshop. My intention is to bring the reader virtually into my shop to generally see how I do the work and share in my thought processes. Although I try to point out how the tools work and the applicable safety considerations, this discussion is not intended to be a text on how to work with tools, nor how to execute operations with the tools shown. I am not providing instruction in woodworking methods. Woodworkers attempting to imitate my methods do so at their own risk.

Click on the images to enlarge them and see the detail

I love this job! Being a custom studio furniture maker means that I get the most interesting work to do! When someone needs something really different, something nobody else can or will build, I am in my element. Not since I was commissioned to make the Rolltop Desk last year have I had such an interesting request as the couple who approached me to build a cabinet that could hold a retractable television ..... and a snuff bottle collection ..... and necklaces and earrings! Now, that's probably something you've never seen before, so the challenge was on!

Shifting into design mode, I visited my clients in their home to talk with them, get an idea of their tastes and expectations, and see what kind of environment the piece would occupy and what the neighboring furniture was like, then I launched into my computer aided design (CAD) program, DesignCAD and made up a few proposals. I try to offer some radically different ideas to help home in on what the clients like and don't like. I'm convinced that everyone, deep down knows what their furniture should be like, and my job is to draw that out so that I can understand and can put ideas into objects. In this case, my client had the requirements clearly in mind, and that was very helpful. After a couple of weeks of back and forth with computer generated renderings burning up the internet connections, we refined the design until we had the perfect piece. The design we came up with has a TV lift mechanism housed in the rear of the case with a portion of the top hinged like a trap door. We located a very solid, simple, and reliable looking lift mechanism called the VersaLift from Wood Technology; the price was right and the fit was perfect for our needs, and it used a radio frequency remote control so we would not have to worry about hitting the little red sensor with the remote's beam, as we all have to do with our TV remotes! Back to the cabinet ... forward of the TV "storage" area is the showcase area behind glass doors. This area has four glass shelves sized for a beautiful collection of snuff bottles. I have to confess that I was not too sure what a snuff bottle was, so if you are culturally challenged like me, a snuff bottle is a small ceramic bottle only a few inches tall. It has a small spatula or spoon that fits into a small hole in the neck of the bottle. According to Wikipedia, snuff bottles became popular in China during the Qing Dynasty (i.e., a long time ago) because smoking tobacco was illegal, but using snuff, or powdered tobacco was fine. OK, that's the end of the history lesson; now onto the most unusual aspect of the cabinet -- it had to house necklaces that could hang and earrings that could easily be selected. What we settled on is two vertical "trays" (think shallow drawers on their sides) that would be installed either side of the showcase area and would pull out on slides. And, I was asked to incorporate the style of legs that I used in my unusual File Cabinet. So, here is what the final design looks like:

After refining the design to make sure all the parts would fit together, which is the real strength of using CAD to design furniture, I started to cut the parts out of my lumber stock. My client really liked the look of sycamore, and it turned out that I had just enough left from a sycamore tree I milled a couple of years ago, but I would have to turn it into veneer. That was fine, since the part of the cabinet the sycamore would be used in, the panel behind the snuff bottles (visible through the glass doors) had to be a dimensionally stable panel. What I mean is that the panel would not be allowed to "move" as humidity changed throughout the year. This might be a good time to digress into a little wood engineering. Wood changes size in width and depth, but not length, as relative humidity changes. I like to say that a board is like a bundle of soda straws where the straws can get fatter if they absorb moisture, but they get no longer. Imagine building a piece of furniture with a 12" wide piece of cherry, plain sawn (meaning it was sawn off the log as essentially a tangential piece, which is the most common type of cut you will see) in my air conditioned workshop in September. The relative humidity in the shop might have been 45-50% all summer and the wood was at 9% moisture content. Now fast forward to February, where the relative humidity in my heated home is 30-35% in the dead of winter, equating to an equilibrium moisture content in the board of about 6.5%. Just that 15% change in relative humidity could cause the 12" wide board to approach 11-15/16" wide. Now, you say, big deal. It's only 1/16"! But I say that if that same board had been glued cross ways to another board such that the grain directions were at 90 degree angles to one another, the change in width would be restricted by the refusal of the mating board to change in length. The force developed by the board wanting to shrink in width would be tremendous, and eventually the glued joint would fail, possibly with an explosive sound (don't ask how I know this; just trust me)! This might explain a crack in a table top where the apron was glued to the top with grains at 90 degree angles. The other possibility is that the entire assembly will distort in shape,

which is a primary reason you sometimes see a dished in table top. So, now I always design so that the wood can move. If I cannot tolerate wood movement for some reason, I use a very stable material such as plywood as a substrate for veneer (or I buy pre-veneered plywood). In this cabinet, all the pieces of the sides, bottom, and top have grain orientations in the same direction so the cabinet will change in depth a little bit seasonally, but will never split or deform. But the panel behind the snuff bottles would give me real problems if it wanted to grow in width, so I am using a very high quality 11-ply core hardwood plywood that I will veneer with sawn sycamore. Which brings me to sawing the sycamore veneer. I used my bandsaw to "resaw" a length of

sycamore, shown in the photo at left. The pieces of sycamore were sawn to a little over 1/8" thick so I could sand them smooth and still have enough meat to later glue to the plywood substrate for the panel. I used my wide horizontal drum sander to thickness sand each piece of sycamore so that the final thickness was about 3/32". Then I set the veneers in a drying rack to stand for a week or more to acclimate to the humidity in the shop. You must understand that when you split open a plank of wood, the center of the wood is probably not going to be the same moisture content as the outside, since it takes time for a piece of lumber to come to an even moisture content throughout and with the humidity changing seasonally the center is always in a catch-up mode. While the veneers were sitting, I brought the plywood up from my storage area, which is humidity controlled, also, and cut the panel ... then set it aside to acclimate to the shop humidity (which is currently a little lower than in my storage area). Letting wood acclimate to a shop is a big deal, and can extend the time required to make a piece of fine furniture. I try to think ahead and keep a good on-hand stock of different species of lumber actually in my shop to lessen the wait times when rough milling lumber for a project.

With the panel parts settling, I focused on the sides and bottom pieces. I make that plural because there are inner and outer sides and the same with the bottoms. This is a complicated piece and is not simply one box -- it's more like a box within a box. These pieces would all be wide and solid cherry, so a single board would not do. I would have to make up wide panels by edge gluing two boards together. The sides would be the most visible, so I chose the finest cherry I had in a thickness that I could

resaw and book match. That means, I sawed a thick board into two thinner boards, opened them up like pages of a book and prepared to glue the edges together. For the bottom and inner parts, part of the areas would not normally be visible so I simply made close matches in grain and color but did not go to the expense of book matching, saving my client some money by using thinner stock. We are still talking quality cherry, here, but thinner wood is less expensive than thicker wood. After rough milling these parts, I also set them aside to acclimate before doing the edge gluing. The photo at right shows the stack of parts on my drying rack, all stickered to allow free air flow.

Next I set to making the back of the cabinet. Again, this part is a big expanse of wood and seasonal growth and shrinkage had to be considered. But another design consideration came into play, here. This cabinet was to house electronic and electrical components. There was the satellite receiver, the TV itself, and the lift mechanism with it's transformer. These components would all need to dissipate heat! Normally we don't think about this because electronic components are out in the open, but if you restrict even a seemingly benign component like the satellite receiver to a small volume, there will be no place for the heat to go and the volume will get hot. This will reduce the life of the component, not to mention the cabinet, itself. So I designed the back to allow free venting. Some makers might simply not use a back at all, but my furniture is always designed to be used in a free standing, middle of the room environment if desired, so a stylish back was a must. Also, the back lends structural stability to the piece. For this piece I designed a vented frame and panel back, much like the back in a previous Entertainment Cabinet. Frame and panel construction accounts for wood movement because the

frame pieces, themselves are relatively narrow and the panels are free to float in grooves. That is why many cabinet doors are constructed as frame and panel. In this piece, the panels are separated by about 1-3/16" to allow lots of air flow while giving the back a pleasing look if the piece is used away from a wall. After selecting the wood and cutting it to the correct dimensions, I set to making the joinery. I decided to use mortise and tenon joints to hold the frame together, so I employed my Domino tenon machine to cut the mortises. This machine cuts a slot for a "wide dowel" called a domino, which is a brand name for a floating tenon. The photo at left shows the machine

with a few mortises, and the photo at right shows the machine in use. After cutting all the mortises, I cut slots for the panels to fit into, and cut

matching tenons on the ends of the panels. Finally, I smoothed all the surfaces with a hand plane and sanded to P400 grit in preparation for a finish. At right I am shown planing a slight chamfer on the edge of one of the back rails, and you can see the mortises and grooves I cut. The photo at left shows the pieces of the back laid out on a table so you can see how it will all fit together.

With the fit satisfactory, I set to applying a finish to the parts of the back. My favorite finish is shellac polish. What is shellac polish, you ask? Well, it is very thin shellac, applied with a cotton pad. I mix my own using shellac flakes. I could make it from store bought shellac in a can, but mixed shellac has a limited shelf life, so I find it more economical to mix my own when needed. The shellac flakes have no shelf life (or at least a very long shelf life - bleached shellac flakes will degrade with time so that they do not fully dissolve). For this project I used blonde and ultra blonde shellac, which is simply shellac that has been dewaxed and bleached to be a pale golden color. I chose blonde and ultra blonde because it goes well with the cherry. Shellac in its raw form is very dark and would color the wood more than I wanted. By the way, shellac is a very natural finish. It is the resin left on tree branches by the "Lac" bug. After the bug leaves its resin, the branches and twigs are cut off the tree and are boiled. The resin separates from the wood and the mixture is spread on flat surfaces for drying. The resulting layer of dried resin is shellac, which is broken up and sold as flakes. The shellac flakes may also be bleached and the natural waxes

may be removed, and that is the kind I like to use -- bleached and dewaxed. All I have to do is dissolve the flakes in denatured alcohol to make what we know of as shellac. My shellac polish is actually 1/2 pound cut shellac, which means 1/2 pound of flakes per gallon of alcohol. I actually mix 1 oz. per half pint to make a jar full. Here you see me in the laboratory mixing a brew of ultra blonde. After mixing the shellac polish, I applied it to each surface and let it dry. It only takes about 10-20 minutes to dry, since it is almost all alcohol. When dry, I sanded again with P400 grit paper, then sanded again to P600 grit. Why didn't I sand to P600 before applying the first coat of polish? It is because any time you apply a finish to wood, there is a potential of raising the grain in the wood. That means that the outer fibers of wood absorb the liquid and expand, making the surface feel rougher than before the finish was applied. So, after the first coat dries and the grain is raised, I simply sand again with P400 to shave off the fibers and get the surface smooth again. P600 would be too fine to effectively shave off the fibers, so I don't waste time with it until after the first coat. Normally I will repeat this process for six or seven coats to get a really nice finish that people love to touch. Only after the finish is right do I glue up the parts.

The glue up of the back was fairly straight forward. First I glued the dominoes into the mortises of the rails, then glued the mortises of the stiles and assembled the entire frame. Clamping the frame to ensure tight joints, I checked for squareness by measuring the diagonals of the frame, as show at right. Although a glue up can really raise a sweat, with all the fitting that must take place before the glue sets up, proper prior planning and laying out in advance all the things needed make the evolution manageable.

With the back complete, I next glued up the parts for the bottom, sides, and sub-sides. You may recall that I had rough milled all the parts and they were settling on the drying rack. Well, now they were ready to edge glue. First, I jointed one edge of each piece to be edge glued, using the jointer. The photo to the left shows the jointer in action. It is a long cast iron beast with a flat top and a rotating drum containing three eight-inch long blades, or knives. A good place to never have your fingers! The jointer gets one face of a board flat and allows you to flatten an adjacent face to a preset angle, usually 90 degrees. After jointing an edge, theoretically

it should fit precisely to another jointed edge. In reality, I often have to follow up machine jointing with a couple of strokes from my hand plane. When the fit is such that there is no gap along the length of the edge and the faces of both halves are in the same plane, it's time

to glue. I used Franklin Titebond III glue for this job, kind of a "super Elmer's glue" because it is easy to work with, is usable straight from the bottle with no mixing of parts (like epoxies), is strong, and has a longer working time than many other glues of its class. I also made the joints without any biscuits or tenons. Some folks like to use biscuits or tenons to strengthen the joint, but in fact if the edges are jointed properly and the glue up clamped with adequate pressure, the strength of the joint will exceed the strength of the wood, itself, so no extra "strengthening" is required. At right you can see me clamping two pieces edge to edge, and to the left you can see both sides clamped and drying.

After gluing the panels together, I let them set on the drying rack some more to relax. In the mean time, I set to work on the legs. As you can see from the rendering (the first photo in this article), the legs are curved. They are constructed with a core of red oak to which laminations of cherry are applied

front and back. The first step was the layout, and again CAD shined. I could print out full size drawings of the legs, tape the pages together (see photo at left), and then cut out the exact shape of the front and side views of the legs. From those "flimsy" templates, I could trace onto poster board and get full size, more robust templates. The photo to the right shows the poster board templates with the flimsy templates to the immediate right. With the poster board templates complete, I could discard the flimsies.

It's now been six days since my last update. I was sidetracked for a few days with the first annual Artisan Craft Conference and Business Institute, held in Roanoke and organized by the Artisans Center of Virginia, or "ACV". At the conference, I gave a presentation on business insurance for crafts people and attended a number of other very worthwhile presentations and panel discussions. The Artisans Center of Virginia is the official state center of craft, located in Waynesboro, VA, and we try to promote craft in a number of ways, including running a gallery in Waynesboro. I am a member of the Board of Trustees for ACV, and I am the editor of The Virginia Artisan, ACV's newsletter, so I'm pretty involved with craft in Virginia. Next time you are driving over the Blue Ridge on I-64, take a few minutes to stop in the gallery and see some of the finest craft art available anywhere! OK, enough advertising; now back to the cabinet!

I was last finishing up the templates for the legs, which were designed as curved red oak cores with cherry laminated fronts and backs. Now it was time to cut out the oak stock and shape it. The first step was to find a substantial red oak plank that could be used for the cores of the legs. I chose a 2 inch thick plank with the idea that I could laminate two thicknesses together to give me enough stock for the legs. Since the legs have a curve, actually in two dimensions, I needed thick stock to work with. So, I set to cutting the plank into manageable pieces, as you can see in the photo at right. I

use my DeWalt sliding compound miter saw for this. This is a pretty neat saw. It has a 12" blade and can be adjusted in just about any way imaginable. It also has a sliding carriage, so it can accommodate wide boards. I mounted it on a portable stand that has wing supports so I can place long planks on it to cut to length. It is accurate, and also portable! And the price was right when I bought it. A very good investment! So, with the plank cut up, I then used my band saw, jointer, and planer to make blanks that could be glued up to the right thickness, and the photo at left shows all four legs glued and clamped. I

had done some preliminary curved cuts, using the templates, to save on lumber, and if you look closely at the photo you can see what I mean. Once the glue had dried, I used the templates to clearly sketch the shapes of the legs, which I then band sawed to rough shape in the side view

dimension, only. The idea was to get the front and back faces of each leg smooth so I could glue 5/16" laminations of cherry to them, then cut the front profiles. Using my stationary belt sander (the photo to the right shows the sander in use for some chair parts I previously made) I smoothed the convex faces on the back of the front legs, and using the jointer I squared up all inside faces and the front faces of the back legs (are you confused, yet?). I also used a spokeshave and oscillating spindle sander to smooth up the concave surfaces on all the legs, shown in the photos at left. Next I would cut the cherry for the laminations.

The cherry laminations for the legs are glued to the front and back of each leg. Think of a sandwich, where cherry is the bread and red oak is the meat and cheese .... and mayo, and lettuce, and tomato ...... hmmmm, that sounds good ... I'll be right back .... .... .... .... OK, now that I've had lunch, let's get back to the cabinet legs. The cherry lamination is

about 5/16" thick. The fronts of the back legs have no curve, so I could cut a full 5/16" thickness and glue it right onto the oak. But the other surfaces are curved and a 5/16" thickness will not easily bend, so I cut strips of cherry half that thickness and laminated two layers together along the curve. The photo to the left shows me thicknessing cherry strips to 5/32", using a flat piece of maple as a base to run the strips through the planer. Those 5/32" pieces bend pretty easily, and using Unibond 800 glue

they hold the curve well. Unibond 800 is a two part glue that forms a rock hard joint that will not "creep". Creep is when a glue gives a little and allows the wood pieces to slightly misalign, and is a characteristic of most carpenter type glues. Normally, creep is not a big deal, but if you are trying to make a lamination hold a curve, you don't want creep. I almost always use Unibond 800 for

curved laminations, unless the curve is slight or there is some mechanical restraint to keep the curve "curved". The glue is a bit messy, so I use gloves when working with it. It's also a bear to clean up the brush and roller I use to apply it, but the extra work is worth the inconvenience when creep can be avoided. The photo to the right shows me clamping a pair of laminates to the curved leg, and the photo to the left shows a pair of legs clamped up for the night while the glue cures.

While I was gluing the leg laminations, which took several days to complete as there were a total of eight gluings, I veneered the sycamore to the plywood for the sub-back that would go in the back of the snuff bottle showcase. You may recall me sawing the veneer pieces earlier, and preparing the plywood substrate. Now I had to edge joint the eight pieces of veneer and glue them to the plywood. First, I edge jointed (made straight

and square) one edge of each of the eight pieces of veneer, as shown in the photo to the right. Then I rip sawed the other edge of each piece to the proper width, which turned out to be 3-5/8". With the edges all fitting tightly together, I next taped them together using veneer tape. This is a thin, water activated tape that holds the edges together while gluing. The tape shrinks as it dries and pulls the pieces together very tightly. The photo at left shows me taping the veneer. While the tape was drying, I did a test run in my vacuum bag press. This tool is a polyethylene bag with a hose attached. The hose goes to a vacuum pump and when the piece to be glued is placed in the bag and the bag is sealed, a vacuum can be

drawn inside the bag that will press the veneer to the substrate with up to 15 psi (usually it's more like 11-12 psi). I made my vacuum pump with parts from Joe Woodworker.com. This guy has a pretty neat web site where he teaches how to build a vacuum pump and how to veneer with a vacuum press. My pump is a venturi pump, where my air compressor supplies compressed air to a small venturi pump that can pull about 24" of vacuum. I need at least 21" to veneer, so I feel pretty good about my setup. The photo to right shows the vacuum bag setup with the sub-back and veneer inside. I bordered the assembly with blue masking tape to prevent glue squeeze-out from getting on the inside of the bag. The tape has an added advantage of making the piece show up in the photo. Experienced vacuum baggers might notice that I am veneering without a platen. A platen is pretty much a top and bottom sandwich of plywood or medium density fiberboard with air passage grooves sawn in the top and bottom faces. The platen holds the assembly flat and helps with the evacuation of air. In this application, the thickness of the substrate was substantial enough to obviate the need for a platen, so I did not use one. By the way, for the glue I used Unibond 800, the same as used for the leg laminations.

The sub-back came out of the vacuum bag the next morning, and the veneer gluing was a success. The first thing I did with the panel was to sand it smooth in the

horizontal drum sander, as shown in the photo to the right. This sander has a cantilevered arm containing a 22 inch long aluminum drum connected to a motor. I wrap sandpaper around the drum, barber pole style, and under the drum is a moving belt that carries the wood through the sander. I can load the drum with different grits of sandpaper, and because the drum is cantilevered, I can actually sand a panel twice the length of the drum using two passes, one pass per half of the panel. I think you get the idea from the photo. This sander makes quick work of flattening a panel, and this veneered panel came out nicely!

By now the legs were all laminated, and so I had to trim the cherry laminate to match the contour of the red oak leg centers. For

this I used a flush cutter with ball bearing guides in my router table. The photo at right shows how this is done, but basically the cutter has guide bearings top and bottom that can ride on a template, in this case the oak cores of the legs, and trim the laminate to the same profile. At this point one side of each of the legs had already been shaped to the final profile, so I could trim the laminate on those sides. The other side of each leg had to be cut on the bandsaw. With this cutting done I could then clean up the saw marks with a spokeshave and then really clean up the curves on the oscillating spindle sander, as shown at left.

I next cut the sides and sub-sides to the exact dimensions. The sub-sides are just like the sides but are located immediately to the right and left of the snuff bottle showcase. The sides, on the other hand, are the most outboard and are on the outside of the pull-out jewelry trays. Since the front of the cabinet is bow shaped, I had to cut the front edges of both the sides and sub-sides to a slight bevel. I used my Saw Stop 10 inch circular saw for this, tilting the blade the correct angle

of bevel as seen in the photo to the left. The Saw Stop is a really great saw. Not only is it hefty, at least half again as massive as my other cabinet saw, but the Saw Stop has a special mechanism that will stop the blade in 5 milli-seconds if it contacts skin. So ...... if the unspeakable happens and my finger accidentally contacts the blade, it will stop before I incur more than a scratch. The inventor of the saw actually tested the mechanism on his own finger (ouch!), and since the saw has been in distribution, there have been many more live tests. It works! Check out the Saw Stop website for the Hot Dog demo, and you can see how the blade stops and drops below the table if a hot dog is pushed into the running blade. No ketchup required for realism, here! The dog gets only a nick. Getting back to the project, after trimming the sides and sub-sides, I trimmed the bottom, sub-bottom and back to fit. The next

big operation was to dry fit everything and drill screw holes and dowel holes, but first I needed to cut dovetails on either side of the front apron. This piece is 7/8" high by 1-1/2" deep and extends between the two sub-sides at the front top. It will serve as the top framing for the doors and give stability to the sub-sides. I cut a simple dovetail in each end of the piece and cut matching sockets in the sub-sides. The trick to cutting a good dovetail is to scribe the cut

lines with a sharp blade and then saw in such a manner as to split the scribed line with the saw to the waste side of the cut. Doing this obviates a lot of chisel trimming. It's not as hard as it sounds to split a scribe with a saw cut if you are using a good saw. I prefer a Japanese pull saw because it has a thin kerf and tends to straighten the blade as it is pulled, much like a string straightens as it is pulled. The photo to the right is a close-up showing the technique. You will have to click on the photo to see the detail in a blow-up. After cutting the sides of the socket, I used a chisel to remove the waste from within, and the photo to the left shows that. Notice that I have a thick maple block to guide my chisel to a 90 degree angle to the work. The block has sandpaper glued to the bottom and it is clamped to the work so it will not shift. I simply align the edge of the block with the scribed mark at the bottom of the socket and chisel half the depth from one side. I flip the work over and complete the chiseling. Nice and clean!

In this project I decided to do all the assembly dry, make sure everything fit correctly, then disassemble and finish the parts, then reassemble. Using stainless steel screws and dowels to assemble the various parts made this possible. The photo to the right shows me drilling a hole for a screw through the bottom of the sub-bottom into the sub-back, and the photo at left shows me drilling a hole for the dowels that attach the back to the sub-sides. This project lent itself well to this type of attaching, and it will make for a very strong piece of furniture. On final assembly I will use glue wherever possible, as well. Also notice that I was very careful to ensure all the parts have grain going in the same direction. The cross grain direction for the sides, sub-sides, bottom, sub-bottom are all front to back, meaning that the carcase will expand and contract in depth seasonally, but there will be nothing to constrain that movement and cause cracking and broken joints.

As I mentioned

earlier, the sub-back is made of plywood so it will hardly move with the seasons, and the back is frame and panel to allow its parts to move without affecting the whole carcase. This is what you get when you commission me to make your furniture - attention to details that will have long lasting benefit and make the furniture last for generations. With the basic parts of the carcase dry assembled, I next dry attached the legs. They are basically attached to the sides, and I used a double dowel joint into the bottom of the carcase and screwed the legs to the sides. They won't be going anywhere! The photo to right shows me aligning the legs for this operation - note that the entire assembly is upside down in the photo. Once the legs were screwed and doweled in place, I uprighted the carcase and proceed to some other operations. The photo to the left shows the upright carcase with clamps used to glue a strongback to the inside of the back. The strongback will carry the weight of the TV lift and the TV. The photo below that shows the inside of the back with the strongback pieces in place. You

might notice in the picture some dark marks on the bookmatched sycamore veneer of the sub-back. This particular sycamore came from a tree that grew alongside Route 360, right at the "S" curve in Heathsville (that's where I live). The tree was very old and surely was used to hold a lot of signs advertising different things, so when I milled the logs I discovered at least seven 20d nails (some were discovered with the blade of my mill --- ouch!) The steel in a nail will leave a black mark as the tannic acid in the wood reacts with the iron oxide on the nail. Well, this is character! This wood has a story!

Before I could disassemble the carcase to finish the parts with shellac polish, I had to take the opportunity to make the trays that would hold the jewelry. So I cut all the parts and began to prepare them for assembly as a kind of vertical drawer, or tray. I decided to use a joinery technique for the corners called a floating tenon. Basically, I cut slots in the ends of the boards into which 1/2" wide oak tenons would fit. Then I glued the tenons to the end of one of the boards, and later would glue the other end of the tenons into the mating board after I had finished the inside surfaces with shellac. The photo to the right shows me gluing a floating tenon into one of the boards. The way the trays would be constructed required some of the mortises to be blind, that is, where the tenons would not be seen. To cut the blind mortises to the exact correct depth, which was critical to these parts, I used a bottom milling bit in the drill press with a shop made

mortising jig. Basically, the bottom milling bit is like a regular drill bit without a point, instead with two flat flutes. This bit can be lowered into the wood and create a flat bottom hole. If the wood is moved from side to side while the bit is lowered, the hole becomes elongated, or like a mortise. The photo to the left shows this in action. The only problem with making a mortise like this is that the mortise has rounded ends. Either the ends have to be squared, or the tenons have to be rounded. In this application, the former was the requirement, so I squared the mortises with a squaring chisel, which is simply a chisel that is "bent" to form a 90 degree angle cutting edge. That way, the chisel squares the mortise as it is driven in, as in the photo to the right. With the mortises all done, I shellaced the inside surfaces of the tray, since

I would not easily be able to sand these smooth after assembly. The photo to the left shows me applying the shellac polish with a cloth. I make shellac polish by dissolving one ounce of shellac flakes in a pint of denatured alcohol. That makes 1/2 lb. cut shellac, or about 1/6 the concentration of shellac you buy in a can at the hardware store. Using shellac flakes has the advantage that flakes have increased shelf life over the mixed shellac and I can make up a batch whenever I need it. Also, the flakes I buy are dewaxed so they make up a more clear solution of shellac.

When applying shellac polish, I normally apply about 6 coats, sanding with 400 to 600 grit sandpaper between coats. The surface obtained is smooth and nice. With the parts shellaced, I next assembled the tray sides and glued them. After the glue dried, I cut a rabbet all the way around the inside edge so I could insert a panel to serve as the back. A rabbet is simply a 'ledge', and I used a bit in the router table called, appropriately, a rabbeting bit. The photo to the right shows me doing this operation. After cutting the rabbet, I measured and cut plywood inserts exactly the size of the rabbeted openings. I used walnut faced plywood for this, as the walnut will look very nice with the cherry and the jewelry will contrast nicely with the walnut. I shellaced both sides of the walnut and then glued the panels into the rabbets. The glued up trays are shown at left. The remaining operation will be to bevel the fronts to match the curve of the front of the cabinet, but this will wait until I fit the trays in the carcase.

A critical operation still had to be done before gluing up the carcase. If you recall, the cabinet would house a satellite receiver and that would go on the bottom shelf. To give the piece a finished look when the satellite receiver was not in use, I would put doors on the front of that cavity. At first I was going to design a sliding door system of six panels, but on further inspection I decided a tambour door arrangement would work and look a lot better. After obtaining approval from my client, I went to work on the

critical aspect of cutting the grooves for the tambours to ride in. One groove would be cut in the top of the bottom, and a mating groove would be cut in the bottom of the sub-bottom. To do this I carefully made a template on which a router with a guide bushing could ride, and using a bottom

mill bit (a straight bit) I cut the grooves. The photo to the right shows the bottom with the groove cut into it. You may notice that the groove goes all the way around to the back of the piece. That was necessary to install the tambours after the carcase was glued up. That also allows maintenance on the tambour if ever that becomes necessary. While the carcase was still in pieces, I took the opportunity to cut the hinge mortises into the

sub-sides, as it is a whole lot easier cutting them with the wood flat on a bench than with the carcase glued up. The two photos at left show me chopping a mortise and paring it flat. With all the machining done, I set to sanding all the carcase pieces, inside and out to P320 grit (which is pretty smooth). I then went through the long process of applying shellac polish to all the parts, inside and out, to the depth of six or seven coats. I sand with finer sandpaper in between coats, up to P600 (really smooth), and frankly I lose track of how many coats of shellac I apply. It's not the number of coats that is important, but the feel of the finish. I know when the finish is optimum because it is silky smooth and further sanding causes a lot of nubs of shellac to form on the sandpaper (also a sign

that the sandpaper should be replaced because the no-clog finish on the grit is wearing off). Also, I finished the pieces

before gluing the carcase so I could sand everywhere and not be left with the telltale unsanded finish in corners characteristic of work sanded after assembly. So I glued the carcase as shown in the clamp-up photo at

left, and then glued on the legs as shown at right. With that work complete I glued the bottom

apron pieces on so the bottom would have a nice vertical curve when viewed from any angle (photo at left). With the carcase all glued up I fit the jewelry trays, an iterative job that takes a while to get

a good fit. Since the trays would slide on ball bearing glides underneath, it was important to get a good vertical fit so the trays would not be sloppy when pulled out. Also, I used magnetic spring latches at the top so there would be no handle - just push in a bit on the front and the tray pops out an inch or so. With the fit nice, I then cut the bevel on the fronts and shellac polished the entire tray outsides. The photos at right show the carcase with the trays pushed in and with one tray out a full travel.

Now it was time to work on the doors. I wanted to get the doors pretty much made before I asked my client to come over and choose a wood for the top. These doors were to be curved, a tricky operation but one I had done before on my Bow Front Showcase and in a large scale version in my Tall

showcase. First, I had to make a standard open mortise and tenon door, but with parts extra thick so I could later cut a curve in them. The photo at left shows me cutting an open mortise in a door rail, and cutting the tenon in the stile was a similar exercise. I also had to

cut muntins and blind mortise and tenons for them to fit in to the door frames. The muntins were cut to overlap each other with a simple lap joint. Finally, before gluing up the door parts I had to cut rabbets for the glass and glazing wedges to fit into. This got a little tricky, because the glass pieces are flat and the door parts were to be curved. To get

the right angle I made a jig to guide a router with a straight cutting bit. Once the jig was made, the door rails were clamped into the jig and I routed. The photos to the right show the jig with a rail installed in it and a shot of the router in action. On the muntins I had to use a

rabbet plane to get the level of the rabbet exactly to the level of the rabbet in the frame. The photo to the left shows the rabbet plane in this operation. When I finally felt that I had all the machining done correctly, I cut the curves on the front and back of each rail and planed a gentle curve on the stiles. Then I sanded and shellac polished all the parts of the doors and then glued them up (photo to right). Finally, I spent quite a while fitting the

doors into the carcase. Since I had made the doors slightly oversized, I gingerly pared and planed a little here and a little there until they fit well (photo at left). Then I cut mortises for the hinges in the outer stiles. Finally, with the doors in place I installed rare earth magnets in a small block glued under the top apron and also to a custom fit board I glued to the top of the sub top to serve as a door stop. These rare earth magnets make a very nice and unobtrusive closure. They simply are countersunk into the wood till the surface of the magnet and it's steel housing are flush with that surface. Then, a steel washer (made for this purpose) is countersunk into the door rail. The counter sinks are only 1/2" in diameter, and I install a small circle of an adhesive leatherette material to hide the metallic look of the magnets and protect them from chipping. The installation is very nice.

There were two major jobs remaining; making the top and making the tambour doors for the satellite receiver bay. The top was to be of Ambrosia maple, which is maple that was attacked by the Ambrosia Beetle, leaving an interesting pattern in the wood. In order to get the width required I would be using two pieces of one-inch thick wood edge glued. But, the TV trap door would have to fit nicely, so I decided to make the glue line right along the line that would delineate the forward edge of the trap door. That way I could cut the door out of the rear piece of wood and close up the saw kerf gap before edge gluing the ends of the back piece to the front piece of the top. To strengthen the glue joint, I cut two Domino slots for each of the

right and left piece glue joint. With the ends glued, I next cut the profile of the bow front on the top, then cut the underside bevel, which would be 1-1/2" wide. The bevel would require a lot of wood to be removed, so to speed up the process I removed as much wood as I could using the table saw. I made an extra high fence, tilted the blade toward the fence so the work would not be pinched between blade and fence, and cut the back of the trap door and the sides of the top this way. The photo at right shows the setup on the table saw. Unfortunately, I could not safely cut all the edge bevels that way, since it is tough to cut a curved edge on the table saw, and the backs of the end pieces were too short to safely cut. So, I reverted to the old standby - hand planing. The photo at left shows me hand planing the back bevel. Note that the trap door is in place so the bevel edge looks continuous.

The tambour doors were pretty straight forward. I made 40 slats, each 5/8" wide (except the two center slats that would get handles - they were 1-1/4" wide) and 1/2" thick. I made them long enough to fit the space for the doors plus allow a 1/4" long tab on each end to engage the slot I had previously cut in the

bottom and sub-bottom. Before I could glue the canvas on the back of the tambour slats I finished them with shellac, as seen on my drying rack in the photo to right. Once the finishing was complete, I made a jig to firmly hold twenty slats at a time for gluing a piece of #12 artists canvas to the back with contact cement. The photo to left shows me applying contact cement to the back of the slats in the jig, and you can see

the piece of canvas waiting to be glued. Once the tambours were glued I removed them from the jig, and the photo at right shows the front and back of the two tambour doors so you can get an idea of how this all works.

All that was left to do was tying in all the parts to the carcase. Attaching the top was a matter of gluing and doweling it with a total of 15 dowels, and then attaching the hinged door with nice brass box hinges. These hinges rotate only 95 degrees so the top does not fall backwards if opened all the way. The TV lift mechanism would open the door as it rose out of the carcase. If you click on this link you can see exactly how this worked in the finished piece. Of course, I had to spend a great deal of time getting the door to fit exactly and getting the top of the piece to be coplaner. Since I

was using Ambrosia Maple for the top, which has characteristic markings, it was imperative that the markings be continuous, without any "jog" from top to door, so my solution was to make the top out of two pieces as described earlier. The finished top looked nice! Another job was to make the pulls for the front doors and the tambour doors. I used walnut for them and shaped the door pulls on my oscillating spindle sander. I used to carve the pulls, but experimenting with the sander showed me that I could do a better job in a fraction of the time using the sander, so that's what I do now. I like to think the "purists" of old would have done the same had they owned such a machine. The photo to right shows this process.

The front doors would not be complete until the glass was installed, so I cut some triangular cross section pieces of cherry to hold the glass

panels into the backs of the doors and carefully drilled screw holes so that some very small screws would ultimately hold the glass in place. These pieces do

the same job that glazing putty does in a sash window in your house (if you have an OLD house that still uses putty glazed windows, that is). Cutting and fitting these pieces is painstaking, but the final look is worth the effort, and replacing a broken piece of glass is a snap. Admittedly, cutting the glass is not my favorite operation - I occasionally end up with a break in the wrong place - but it makes much more sense for me to custom cut the glass than to make templates and get a glass place to do the job. The photo at left shows the cutting operation, and the photo at right shows the inside of a finished door. Installation of curved doors presents some challenge to get the fit as perfect as possible, but the look is very nice!

Remember the trays to hold the necklaces and earrings? I still had to make the holders to fit inside the trays. For the necklaces, we chose 12 nice brass pegs to hang them from, and for the earrings I borrowed a few of my client's stock to design a holder that would work well. The end result was room for 64 pairs of earrings. The photos to right and left show the finished trays.

The last step was to install the TV lift and a power strip to plug all the wires into, and apply a coat of Briwax, my favorite wax. Here are some photos of the finished cabinet.