Category Archives: Design


Blanket Chest – Topping It Off

As I mentioned previously, the top of the blanket chest will be made from red oak, to match the rest of the furniture in the bedrooom.  Red oak is not one of my favorite woods – far from it.  In fact, it would be near the bottom of the list of woods I like to use.  Its open grain and proneness to tear-out make it frustrating to work.  Unfortunately, it’s one of the more common woods in our area, so I often have to just set my jaw and plow ahead.

Due to surface flaws, I ended up having to plane my stock closer to 3/4″ than the 7/8″ that I wanted.  This wasn’t a problem, though.  The lid stands alone as a design element, and edge shaping makes the thickness harder to discern anyway.  I won’t bore you with the details of gluing up the panel from the individual boards – there are plenty of tutorials out there on that.  I planned on a 5/8″ overhang on all sides, making it slightly smaller than the size of the base.  It has been my experience that this generally gives a more balanced proportion than making top and bottom the same size, which tends to look a bit top-heavy.

Blanket Chest 41

The overall panel width was just a bit too wide for my crosscut sled, so I had to find an alternative means to crosscut to final length.  While some would use a jig with a circular saw or something similar, I chose a more direct approach.  I drew a line, and cut to it with a 12pt. panel saw.  A quick cut, followed by clean-up with a jack plane, and I was through by the time most people could have found wood for the jig.  Even if you’re not a serious hand-tool woodworker, the ability to accurately cross-cut to a line with a handsaw is a skill that should be in everyone’s repertoire.

Blanket Chest 43

There are lots of different router bits on the market, many with exorbitant price tabs, but following Matthew Bickford’s blog has gotten me to thinking of complex molding shapes in terms of combinations of hollows and rounds.  Even if you can’t afford a set of matching hollow and round planes, you can apply a lot of the principles with a few simple router bits and an edge guide. Not only can you duplicate many existing bits’ profiles, but you can customize things to get the exact profile you want.

My wife wanted a molded edge treatment on the lid, so I played with some options till I came up with something she liked.  Her favorite was a combination of a small cove with a larger roundover.  I started by cutting a sample, then she would take a pencil and sketch in modifications on the end of the board until she was pleased with the outcome.  In the photo above, I started with a cove and roundover combination on the edge, and she suggested changes.  She finally settled on the innermost style.  The roundover has been brought up and in to make it a larger part of the overall profile, while the cove was made a bit smaller to act as more of an accent.  To achieve this, I switched from a 3/8″ to a 1/2″ roundover bit, and cut with it low enough that the guide bearing actually hung off below the edge of the wood to yield only a part of the curved shape.

Blanket Chest 44

I’ve used this D-handled Porter Cable for years, and it’s a favorite of mine.  The addition of a good-quality edge guide allows you to place individual design elements exactly where you want them, instead of being limited by a bit profile.  In the photo above, the basic roundover cut is being “lowered” into place, and then the cove will be cut with a small core box bit  where you see the large shoulder above the roundover.  As mentioned above, the fence of the edge guide lets me run the bit without registering the bearing on the surface, and lets me use partial arcs as part of my design.

Notice also in the photo above the presence of an offset base.  I cannot recommend one of these highly enough.  One of the biggest concerns when doing edge treatments with a handheld router is tipping.  I don’t care how good you are, the darn things seem to want to tip and ruin your work.  Adding a heavy edge guide like the one I have increases the odds of that past the point of acceptability.  The offset base lets you hold the router firmly to the work, so you concentrate on other things like keeping the fence snug to the wood, and proper entry and exit from the cut.  Buy one or make one, but by all means, get one.

Blanket Chest 44

After the edge was routed, I put the lid in place on the body for a look.  All in all, not bad!  All that remains is to paint the body and stain and finish the top.  It won’t be long now!  Stay tuned!


Blanket Chest – A Firm Foundation

The base of the chest was intended from the start to be simple.  My wife didn’t want any feet, openings, or anything of the kind.  She just wanted a simple base or skirt of the type you’d find on a tool chest.  And what mama wants, mama gets.  Besides, what could be easier?

Having read Christopher Schwarz’s The Anarchist’s Tool Chest, I chose to dovetail the base, rather than using miter joints.  This chest would sit on carpet at the foot of the bed, and be dragged back and forth when vacuuming.  Due to this, I felt the extra strength of dovetails would be worth the trouble.  There’s no point in going into the details of cutting through-dovetails – that has been covered by others ad nauseum.  I will, however, cover my technique for getting tight-fitting wrap-arounds with dovetails in my next article.

Blanket Chest 36

I decided on cutting my base 3 1/2″ wide.  This was done by the scientific approach of laying different widths of wood up against the bottom of the chest and, with my wife’s input, deciding which one looked best.  Once the dovetails were cut, the same approach was used to determine how far the base should overlap the bottom of the chest.  We settled on a 2 7/8″ reveal before applying the cove molding.  Yes, that’s almost the entire width of the frame.  However, there are still the bottom boards that project below the frame to be taken into consideration.

Blanket Chest 37

I cut internal supports for the chest from whatever plywood was handy around the shop, and believe me, my pile is as bad as anyone’s.  After ripping to width, they were glued and brad-nailed into place.  As you can see, there’s still an adequate amount of recess to make everything secure.

Blanket Chest 38

Then, the dovetails were glued together around the chest, and the base was attached to the bottom boards with pocket screws.  As we know, in cabinet construction, there’s primary wood, and secondary wood.  But, as the above picture shows, there’s also tertiary wood.  This is one case where your ugliest plywood is perfectly acceptable.

Blanket Chest 39

Here you can see the base fully assembled with the dovetails planed down.  The construction, as it stands, is perfectly acceptable.  However, there’s a certain starkness to it, as though something is missing.

Blanket Chest 40

That’s where the molding comes in.  It adds that needed transition, and harmonizes with the smaller cove molding around the panels.  If I had been using a clear finish, I would have used 23 gauge pins.  However, a painted finish allowed the use of spackling compound, so I stuck with the bigger brad nails and glue.

That finishes the body of the blanket chest!  Now, it’s time to turn my attention back to the lid.  Stay tuned!

Blanket Chest – Fitting the Floor

Blanket Chest 34

Adding the floor was one of the simpler parts of the process.  All of the boards had already been cut to length, so it was just a matter of layout.

I started by finding the center of the blanket chest, and placing the edge of one board adjacent to it.  Then, with the board square across the opening and the bead facing down (remember, down is up here), the board was secured in place with one screw on each end.  If the screws were the sole means of support for the bottom, I would have used two.  However, the entire chest will be resting on a lip on the inside of the base which will provide the actual support, making one screw sufficient.

The most important thing was even spacing of the boards.  Not only do the boards require a small gap for expansion and contraction, but the width of the space should give the appearance that the bead is centered between two equally-spaced gaps.  After playing with various items, I found that 18-gauge brad nails were just about perfect.  On reflection, I should have planned ahead when I made my scratch stock, and made sure the gap it created matched up perfectly with a spacer ahead of time.

Blanket Chest 35

In the end, when the chest was turned over, the spacing was fine.  However, the groove didn’t provide the same shadow as the gap, giving a less-than-equal appearance.  This surprised me, since they had looked much more equal when laid out on the table.  I can only attribute this to the different way that light plays off the inside of the chest.  It’s not a bad look, but not what I intended.  The only consolation is that the bottom will be covered with blankets.  That’s not much consolation to a woodworker, but it’s not worth doing over, so I’ll take what I can get.

Blanket Chest – Dimensioning the Stock

Yes, believe it or not, I’m back on the blanket chest!  After taking a couple of side jaunts to the Herman-style sawbench and the Krenov-style sawhorses (sawbents), I’ll bet you thought I had forgotten about it.  Don’t worry, there was no chance of that (my wife wouldn’t let me if I wanted to).

Blanket Chest 1

When we last left this project, the rough poplar lumber had been skip-planed, stickered, and allowed to acclimate to the shop.  Now, weeks later, it was time to stand it all up and take a good look at it.  At times like these, it’s good to have a cavernous building with a loft to lean your wood against.  This allows me to inspect it all at once, and decide on how to break it down into rough pieces.

In the last blanket chest post, I had established the rough pieces that I would need for the project.  Now, it was a matter of looking at the widths of the individual boards, and figuring the best way to cut them to the required rough parts.  There is more art to this than science, but the process was eased by the fact that everything would be painted.  Also, as you can see, the boards were all very clear of defects, though with considerable color variation.

For this phase, I picked out just the pieces for the frames.  The primary goal was to get two 3″-wide frame components from the width of each board.  All components were effectively matched pairs, so rough crosscutting would yield a pair of matching components with each pass, reducing waste.  Since all of the boards were over 6″ wide, this was a matter of picking the narrower boards for the frame components, and saving the wider boards for the panels.

Blanket Chest 2

After crosscutting to rough and in some cases generous length, I ripped the wide boards to a rough width of 3 1/2″.  After that, it was back through the planer for a final thickness of 7/8″.  Why rip before planing?  This approach makes it easier to remove any bow from the stock, as the narrower boards will flex less in the planer, in many cases eliminating the need to hand-plane a flat face first.  Also, less stock needs to be removed to produce a flat surface, as shown in the diagram in the grooving plane series.  From this point on, even though I didn’t sticker them, I kept the stock stored on-edge on the outfeed table to equalize moisture exchange on both sides.

After this, all stock was ripped to 3″ in width, ensuring good, clean edges, and then cut to final length.  The last step was to cut four of the stiles to a width of 2 1/2″. These pieces will have the 3/8″ tongue portion of the tongue-and-groove joints that will join the frames together at the corners, and have to be narrower to give the appearance of a 3″ width when the two stiles are glued together.

Corner Detail

The math is:

3″ total width – 7/8″ for grooved stile thickness = 2 1/8″

2 1/8″ + 3/8″ tongue length = 2 1/2″

Therefore, when the 3/8″ tongue is inserted in its matching groove, the 2 1/8″ visible width, combined with the 7/8″ thickness of the front stile, will yield a total visible surface of 3″, matching the width of the front stile.

  • Blanket Chest 3

The frame stock is shown in the photo above and, from the left, consists of:

  1. Four long rails for the sides.
  2. Four short rails for the ends.
  3. Four wide corner stiles for the sides.
  4. Four narrow corner stiles for the ends.
  5. Four inside stiles for the sides.

The next step will be to start the joinery.  Stay tuned.

Blanket Chest Design – Mix and Match Moldings

Well, I’ve gotten side-tracked by life, and have yet to make it to the lumber yard.  Not to worry, there are still things to think about in the meantime.  One of the biggest design elements left to work out is the molding on the edge of the panel frames.  One of the most widely-accepted ways of doing this is with a matched set of door-making bits for the router table.  However, I’m not personally inclined to this approach.  It locks you into one molding style for each set of bits purchased (and they ain’t cheap!), and setup can be very tedious.  That’s fine if you have a commercial shop, but it just doesn’t suit my temperament and penchant for hybrid woodworking.  Besides, I believe a traditional mortise-and-tenon joint is much stronger overall.

Applied Molding Detail

My plan is to use applied moldings.  I mentioned in a previous post that I hope to be able to wind up with finished stock that is 7/8″ thick, rather than the standard 3/4″.  This will allow me to place a 1/4″ groove 1/4″ from the back face, and have 3/8″ thickness on the front face.  This will give me plenty of room to use custom made applied moldings, and even have a small rabbet shoulder at the top edge for effect if I desire.  The mortise-and-tenon joints will be centered in the grooves.

Panel Trim 2

Applied moldings have several advantages over rail-and-stile bit sets.  First, you can custom-craft your edge effect on the fly.  In the picture above, you can see four test pieces that I made just using bits laying around the router table.  The combinations are limited only by your imagination and budget for bits.  Router bits of this size, are far cheaper than their larger cousins.  This means more bang for the same buck.  Secondly, the applied pieces can, if used carefully, cover any gaps in your casework, and make for a smoother look.  Last, but certainly not least, a mistake in joinery won’t cost you nearly as much here.  If you mess up a molding, simply make another to fit.  You should, by the way, cut extras – it’s not hard.

Cutting Molding

The important thing to remember is that small moldings are cut from big stock.  Use your router bit to cut the desired profile in the corner of a larger piece of wood.  Then, make cuts with your tablesaw to release the molding from the stock.  I’ve found that I get better results with stock that is slightly longer than the finished pieces.  This seems to yield better results than by using long stock and cutting to length, probably due to the fact that the shorter pieces are easier to hold against the fence consistently.

The problem is, getting my wife to pick a molding!  I’ve shown her the four samples, and told her that I can get bits for other profiles if she prefers.  Well, I’ve still got to get the wood and let it acclimate, so the need for a decision is some distance off.  Until then, the samples sit in the dining room, waiting for her to pick a winner.  When will that be?  The world may never know!

Panel Trim 1

Blanket Chest Design – Lumber Calculations.

Now that the basic dimensions have been established, it’s time to determine how much lumber I need to buy for this build.  At this point, I am calculating for the basic frame-and-panel box and the bottom.  The wood will be poplar, and a good 4/4 rough.  I would actually like it a little thicker, so that I can plane down to a finished thickness of 7/8″.

The overall dimensions of the chest are 48″ long x 22″ high x 18″ deep.  Since the rails and stiles will be 3″ wide, I will need 3″+ wood for that.  I say 3″+ because I’ll be ripping to 3″ in the end, so anything over that will be waste.  Due to this, I’m not thinking so much in board feet(bf) as liner feet(lf).  I’m planning on 2″ tenons on all joints.  For the rails and stiles, the calculations are:

Front & back rails:  4 x 46″
Side rails: 4 x 16″
Corner stiles: 8 x 22″
Inside stiles: 4 x 20″
Total Wood:  504″, or 42 lf.  Allowing for waste, this will require four 8′ boards 3″+ wide.

For the panels, there is a bit more variability.  Panels will be 12 1/2″ wide and 16 1/2″ tall.  The 1/2″ allows for tongues on all four sides.  For the sake of flexibility and to allow for any alignment issues, I’m calling the panel pieces 18″ long.  For the 8 panels, this comes to a total of 12 bd ft.  If I use 3″+ stock, it will take 48 lf, or six 8′ boards.  If, on the other hand, I use 4″+ boards, it will require 36 lf, which will require five 8′ boards.

For the bottom, I’m planning on using tongue-and groove boards running across the width of the carcase.  These will be 18″ long, and 3″ or 4″ wide.  That means that the bottom will require 6 bd ft.  If I use 3″+ stock, that will require 16 pieces, for a total of 24 lf, or three 8′ boards.  If I use 4″+ stock, it will require 12 pieces, for a total of 18 lf, which would still require 3 boards.  The best bet is to stick with 3″+ boards.

So, to sum up, I will need 13 8′ boards 3″+ wide, or seven 8′ boards 3″+ wide and five 8′ boards 4″+ wide.

Now, it’s just a matter of going to the lumberyard.

Blanket Chest Design – Wood Moves.

Warped Panel

Wood is always full of surprises.  As you may recall, I had glued up some test panels to work on my raised panel design for the blanket chest.  After settling on one, I left the other one on my bench.  When I went back a couple of days later, I was greeted by the above panel.

Yes, we all know that wood moves, and yes, we all know to let wood acclimate to our shop before using it.  The thing is, this particular wood had been in the shop for almost two years!  It was some leftover knotty SPF that I hadn’t found a subsequent use for, and it had been through two full climactic cycles supported on a rack.  If any wood could be called acclimated, this was it.  Furthermore, this was a glue-up of three separate boards that were subsequently flattened.

Nevertheless, here it lay on the bench, mocking me.  As you can see from my hi-tech aluminum angle winding stick, the bow is significant.  If I were indeed planning to use it as a panel in the blanket chest, that plan would have to be abandoned.  Obviously, internal stresses were released when the panel was shaped, but I’m not yet sure what or why.  Was the fact that more wood was removed on the other side what allowed it to curve this way?  If it were simply differential drying, I would expect it to cup towards the raised-panel side, since more end grain was exposed there.  I’m tending towards the physical weakening theory, but that’s all it is – a theory.  If any of you can enlighten me, please do.

In any event, this is an unsettling reminder of the fickle nature of the material we work.  When I buy my wood next week, I need to make allowances for such possible panel disasters.  If nothing goes wrong, it never hurts to have spare poplar laying about.

Blanket Chest Design – Panel Raising

Panel Design 1

What on earth is going on with all those clamps?  Simple, panel design.  “Wait a minute,” you say, “I thought design was done with pencil, paper, ruler – things like that.”  A great deal of it is.  However, sometimes nothing works like making an actual prototype.

There are a lot of variables in a tablesawed raised panel’s design.  How thick will the panel be?  How far should I tilt the saw blade?  How much reveal do I need?  How will I treat the back?  Should I give special consideration to the edge that fits the groove in the frame?

I’m sure there are plenty of ideas out there about what makes a good panel design, and I don’t doubt that there are some “rules” on the subject.  However, if you’ve followed this blog for any length of time, you know that I have a tendency to forge my own path on things.  To me, the best way was to make some test panels and try things for myself.

Once the panels were dry and cleaned up, I set up my new panel-raising jig and started making sawdust.  I began with a 5 degree tilt and a 1″ reveal (allowing 1/4″ for the groove and 1/4″ for the trim, the actual depth was 1 1/2″).  I then made adjustments through a very systematic process:

  1. Cut panel.
  2. Show panel to wife (better or worse than the last?)
  3. Go back to shop, make adjustment, and cut another panel.

Panel Design 2

Going along in this manner, we finally arrived at something that looked right:

Panel Design 3

The winner had an 8 degree bevel, and a reveal of 1 3/4″.  This worked out to be exactly half the width of the frame if the trim is included.  I don’t know if this was coincidental, or if this is one of those “rules”, but it worked for us.  I did a slight stub tenon on the one above to play with fit into the frame, but this is easily covered by trim.  I’m still working on that part.

Next time, I’ll go into the oft-mentioned trim, plus my ideas for the frame-panel relationship.  Stay tuned!

Blanket Chest Design – Panel Proportions

Now that the basic construction principles had been decided upon, the next question was one of sizes of the various components.

Overall dimensions for the blanket chest were actually the easiest to establish.  I took a tape measure and called my wife over to the intended location at the foot of the king-size bed.  By extending the tape measure and asking, “Is this long enough?  How about this?  You’re sure you want it that long?  Okay, okay!”, we arrived at a length of 48″  Remember, this is momma’s chest.  We decided on a height of 22″ – a bit tall for comfortable seating for shorter people, but my wife wanted the increased storage, and pointed out that it wouldn’t be a regular “sitting seat”.  Likewise, the width was set at 18″.  I had thought that 15″ would give a better fit at the foot of the bed, but, again, my wife wanted more room for storage.

I’m not a SketchUp user, but have ample experience with traditional drafting.  Using these skills, I started proportioning the components of the chest body.  After fiddling with rail and stile widths, I finally settled on 3″ wide rails and stiles as giving the best look for the overall proportions of the front of the chest.  After reviewing mockups, we decided on a three-panel front and back, and a single panel on the sides.  I started a final drawing of the front to establish panel dimensions, and this is where the fun began.

I started by simply dividing the front into three parts, and centered the stiles on these marks.  When I finished the drawing, something didn’t look right:

Panels Divided Outside
The center panel was wider than the two outside ones!  How had that happened?  The areas were equally divided, or so it seemed from the divisions at the bottom, but the center panel was wider.  This was not the look I wanted.  After a few minutes of head scratching, I tried a different strategy, this time starting my divisions at the inside edge of the outer stiles:

Panels Divided Inside

This time, the center panel was smaller than the two outside panels!  Aha!  I thought I had it figured out, so I tried the remaining logical alternative:  I made my divisions starting at the middle of the outer stiles:

Panels Divided Middle

Hey!  It worked!  All the spacings are equal.  So, I concluded, the secret to even spacings is to start your divisions from the middle of the outer stiles, right?

Not necessarily.  This will work if the outer stiles are the same width as the inner ones.  However, this truism breaks down if the width is different.  If, for example, you have wider post-type stiles, or just want the look of narrower inner stiles, using the above conclusion will get you into trouble.

In fact, the divisions should be made starting at a distance of one-half the inner stile width out from the inner edge of the outer stile.  For example, if the inner stiles are 3″ wide (as above), the equal divisions would start 1 1/2″ left of the inner edge of the left-hand stile, and end 1 1/2″ right of the inner edge of the right-hand stile.  In this case, this corresponds to the center, but the actual width of the outer stiles is irrelevant.

I hope that’s not too confusing.  It’s one of those useful things that I don’t ever remember seeing in print before.  Play with it a little, and I think you’ll get the gist of it.  If not, leave a comment here, or contact me at, and I’ll try to clarify things.

Blanket Chest Design – Turning the Corner

I’m still playing with the overall size of the chest, but have another related item that must be addressed – the corners.  Since my wife wants frame and raised-panel construction, there are a few options available on joining the front, back, and sides together.

The first option is using a corner post as the end stile of the frame-and-panel assembly.  In the variant shown above, the rails are set back some from the front edge, so that the post stands somewhat proud of the rest of the frame.  This is often used when the post will act as a foot for the chest.  Again, pardon my poor drawing – I just haven’t gotten around to SketchUp.  The apparent stub tenons are actually the exposed part of haunched mortise-and-tenons as shown below:

Stub Tenon
Keep in mind, however, that my wife wants a simple raised plinth around the base of the chest.  Projecting corner posts make this difficult to achieve.  A variant of this approach moves the rails out to sit flush with the face of the corner post:


This gives the appearance of a continuous frame-and-panel structure, rather than one terminated with a larger post.  This makes the addition of a plinth around the bottom much simpler, since there is no need to work around the projecting profile of the corner post.  There is, however, one drawback to this approach:  A significant amount of post projects into the interior of the chest.  If the chest is to be a simple open box, this is basically irrelevant.  However, if you plan to incorporate a till or lift-out drawers, these posts will certainly get in the way.

There remains yet another method of construction that doesn’t require a corner post.  In the above example, all four sides are completely constructed as independent frame-and-panel elements, and then glued together at the corners to form the basic box of the chest.  While first impressions are of a weaker joint, bear in mind that this is long-grain to long-grain, so it should be stronger than the wood itself.  The addition of full-length stub tenons can add a great deal of physical strength to the joint as well.  The only thing to keep in mind is that the end-panel stiles have to be made narrower by the thickness of the front and back panel stiles so that the look is symmetrical from any direction.

Of course, the visibility of this sort of joint would cause problems on a clear-finished piece.  I have seen methods for hiding such an overlap, but am uneasy with the idea.  However, painting makes it a perfectly viable alternative.  Since my wife finally decided on a painted finish, and that lift-out trays are a possibility, this third method seems to be my best option.  One thing about painted finishes, they certainly allow for a different set of joinery options.

The next step will be to finalize the basic carcase dimensions.  In other words, my wife has to commit to how big she wants it to be.  THAT will be the hardest part.  Stay tuned!