I am building a new workbench for my in-office, rainy day woodworking. It will replace one that served me well, but has some shortcomings. Plus, I really was driven to try the classic leg to top Roubo joinery.
First up was milling stock for the legs. I used the 8/4 cherry from Bell Forest Products for the entire project. The bench will use the better part of three of the six boards I received (100 b.f.). The legs are glued up to created 3.5 x 2.75 inch components.
This simple, but robust mini-bench will not have any vise, but I can use my Moxon style vise when that form of work holding is required. Mostly I anticipate using dogs, holdfasts, battens and wedges.
This was a fairly early project of mine. I am slowly bringing images over to the blog site. This was fun to review. My methods have changed considerably over the years, however, the basic structure is very much similar to my current approach to cabinet construction. Plenty of interlocking joinery to amuse this woodworker.
Around 2010 I built a sideboard using highly figured Bubinga. I was still in my experimenting with finishes stage. The top was too shiny for my taste, but I left it that way for a long time. Finally, I decided to refinish the top more to my liking.
Removed the top and sanded with 50 grit paper and proceeded through grits to 220. The original top also showed planer snipe at one end which was only visible in certain lighting angles. The sanding also removed the snipe. Thank you, very much.
Fresh matte finish and snipe removed
Finished with usual varnish/BLO/mineral spirits wipe on/wipe off procedure. Besides being more to my taste, I removed the wince that would cross my face every time I looked at that piece.
In response to Shawn’s query I looked at the original photo gallery (and the actual cabinet). Flipping through the images I see a number of processes that no longer fit my style. The hollow chisel mortiser is gone. Replaced that process with either uncut spiral bit in the router (with edge guide) or by hand with mortising chisel. I do not use a pattern router bit on curves any more. These days I’ll cut close to the line on band saw (as before) and then shape with planes, spokeshaves, rasps and sandpaper.
I do not ever expect to use the hidden Soss hinges again. They caused all sorts of issues. My solutions created rather weak stiles on the hinge sides as I had to round them over to gain clearance. These days I would use knife hinges.
Side pieces surface glued. Note figure on outside piece, more common figure on inner
Shawn asked of rationale for not using secondary wood for non-visible portions such as the case top (buried below the show top). I had enough of the fabulous Bubinga for the show top, sides, drawer fronts and door panels. “Regular” (but still nicely colored and patterned) material for the rest. Pure indulgence. For instance, the sides are built to 1-1/2 inches thick from the figured material on outside and regular, but lovely material, on the inside of the sandwich. That is hidden below the top (above) and the legs (front view), but it was fun to do it that way. The inside surface is visible when you have the door(s) open.
Interestingly the hardwood dealer (Soboba) charged the same for the highly figured material as the regular Bubinga. When I saw it in his shop, I immediately got out my checkbook without a specific use in mind.
I used Philippine Mahogany for top row of drawer runners (aka web frame). Also used P.M. for drawer sides and bottoms as well as the back ship lapped back slats. Legs of Padauk.
Small table of Honduras Mahogany salvaged from Belize rivers by Greener Lumber, LLC.
A small table with floating top is one of the few designs I have built several times. The heights vary and some small tweaks in other dimensions to suit material on hand. This particular version was made for a Michigan friend.
Mortises were cut on legs with 3/8-inch spiral uncut router bit and edge guide. Tenons created to fit using dado stack on the table saw before shaping. At this time the ship lap notches were cut for the overlapping stretchers.
The shapes at top of legs was done in several steps. Rough miters (45 degrees) were cut with backsaw, then taken close to layout lines with chisel and final smoothing with sanding block(s). The leg bottoms received a small bevel with block plane. Stretcher curves cut with bandsaw and cleaned up with spokeshaves and sandpaper.
The one-piece top (roughly 14 x 16 inches) was too wide for my jointer, so I used a No. 7 jointer plane to create one side capable of being stable as feed through the drum sander. There was a large missing chunk on one edge. I kept that as part of the table’s character, but did clean up some of the punk material with carving gouges. A Buhl-diamond was carved on the underside with a V-gouge.
After final sanding and easing of edges my usual oil-varnish blend was applied in multiple wipe on, wet sanded and rubbed out coats. Perhaps eight coats total.
First finish coat going on the underside of top. Note punk area and the Buhl-diamond.
The legs, stretchers and top were packed in a suitcase and taken to Michigan for final assembly and glue-up. Hide glue was used for the mortise and tenons and four screws driven up through the upper stretchers into the top hold things together.
Working outside of my usual environment muddled my brain and I mixed up the stretchers. I did not catch this major mistake until after the glue had set up. Knocking it apart split one stretcher in half. So before doing a rebuild I had to glue that together. Not a perfect fit, but hopefully, it will hold up for at least one generation.
It was delivered to my friend on the streets of Detroit near the Wayne State University campus. I trust it is enjoying its new home in the big city.
The Penny Collector poster nestled into its new home.
The Penny Collector poster is ready for hanging in my office. I carved a simple Buhl-diamond on the backside of the top rail with V-gouge. Then finished with multiple, thin coats of Tru-Oil finish. I may take the sheen down a bit after it is fully cured with light 0000 steel wool buffing or I may leave it as is. Time will tell. I found out about Tru-Oil from the BenchCrafted blog [http://benchcrafted.blogspot.com/2015/01/why-you-should-be-using-tru-oil.html]. Jameel’s blog and BenchCrafted are at the top of my list for insights and inspiration.
Cherry plywood for backer
I wanted to make the backer from material on hand. Made one using left-over teak plywood. Liked the look (not that it will matter facing the wall), but the board had severe cup which would stress the hardware holding it in place. Oh, btw, I used two figure-eight connectors for that task.
Next I dug out some cherry plywood left over from one earliest projects (before I swore off plywood). The figure is not particularly pleasing, but it will amuse the spiders as they spin their webs in the dark.
The glass was cut to size (12 x 18) at my local glass shop and placed along with the poster and backer into position. Then it was photo time.
If you missed part one of this short series, this letterpress poster was from Carrie Elkin [http://carrieelkin.com] as part of her Kickstarter campaign for THE PENNY COLLECTOR.
We left the last post with a destroyed lower rail. Before attempting a patch, I felt it necessary to stabilize that rail from the back side. Flowed lots of West Systems epoxy into the rotted areas, but did not attempt to level the surface. Once cured I sanded any epoxy above the rail’s plane.
Drawing rail shapes
I drew a curve for the upper rail as well as angled ends of each rail then blended those with rounded corners. Cut the angles and curves at the bandsaw and cleaned up with rasps and sanding. I also drilled a shallow hole which will have a shiny penny placed once everything is complete.
Dry fit test. Shows hole for penny and the destruction to the right side of the lower rail.
Here you can see the consequence of the destruction. My material supply was pretty limited but I found a piece to create a full-width patch. After gluing the frame I routed a shallow rabbet to capture the patch. When gluing the patch in place I did not notice it had slipped away from the cavity. I used liquid hide glue so was able to apply water and prying to remove the patch.
Patch number two
However, the fragile patch did not survive. Next I made a patch of two separate pieces which matched each other pretty well, though not as well at the ledge as my first attempt. Glued those into place while paying better attention this time. The patch was sanded level with the ledge of original material and edges cleaned up.
After a final sanding it was time to begin finishing. Jameel of BenchCrafted had blogged of Tru-Oil a while back. I tried it for one of my Esherick-style music stands and liked the look. No sure I would do a large piece that way. But this would be a good opportunity give it another look.
A 1-3/4 inch thick hunk of chestnut being trimmed to “straighten” the grain lines.
A picture/poster frame can be a nice one-day or weekend project. Four sticks of wood with a bit of joinery. No big deal. Or, you can work like me and enjoy weeks of fun and lessons learned the hard way.
For my support of Austin-based singer/songwriter Carrie Elkin’s Kickstarter campaign I received some fun swag. In addition to her touch-your-soul music, there was a 12 x 18 letterpress poster which deserves a nice frame. Check out THE PENNY COLLECTOR if you are interested (hint: you should be).
Parts milled (jointed and planed) four-square but a bit over-sized. This allowed me to decide on widths of each element.
For this project I used the last remains of the large Chestnut slab I bought from a Maryland hardwood dealer several years ago. This would be tricky with some rotted areas to contend with. I selected half-blind dovetail or maybe it should be called dovetailed ship-lapped for the joinery. The stiles were identical while the top and bottom rails would each have different thicknesses and widths. The stiles came in at ~1/2 inch thick, top rail ~3/4 inch and the bottom ~5/8 inch.
The four sticks were rough cut from the slab at the band saw, then one face and one edge made true at the jointer. Power planing took the pieces to final thickness. All sticks were over-sized at this point. Rabbets were routed to receive the glass, poster and backer. Stiles were through cuts and the rails stopped, then squared with chisel.
Dovetails cut on stiles. A small ledge was created on the upper face to create a shoulder.
Dovetails were marked at each end of the stiles. Shoulders yield 18 inches between the upper and lower rail rabbets and 12 inches between the stile rabbets. A 1/32 inch ledge was created at the router table, then the dovetails cut at the band saw. Final clean up with chisel to create clean shoulders. In the photo you can see the consequence of making through rabbets. The tail is flush to the socket so should be fine. However, that small triangular area is fragile and broke off on two of the joints. Not a big deal as the contact at intersection is intact. However, as I was cutting these (and the sockets) freehand I did not need to keep them to standard shape. Small lesson learned (more serious lessons to follow).
Ready to begin chisel paring to the knife lines. What you do not see in this image, but can tell if you are familiar with the Roubo wagon vise, is the cavity underneath where I am about to cut with the chisel. Yes, I cut right through to the face with major blow out.
Sockets were first cut with a trim router and straight bit. I stayed away from the knife lines. The router makes a nice flat reference depth (unless you tip the silly router as you work). The top rail went well and fit nicely with just a bit of chisel finessing. The bottom rail began as a minor problem and then escalated. The underside had considerable rot, which I was aware of. I was not, however, aware of how deep and extensive that rot was. Then the big lesson. After routing the base depth, I began to pare to the knife lines and KAPOW! Chisel went right through the face side. Why, you might ask. Well the rail was directly above a cavity in the workbench. That cavity is for the wagon vise block to move in and out. I saved the two large and one small busted pieces thinking I might be able to repair. Later I will discover the puzzle would not go together.
Material was selected for the legs and roughly cut at the band saw. At this stage I am looking to create reference surfaces for the legs while leaving as much material as possible for future shaping.
One of the front leg sticks had punky area at the bottom which I hoped would be outside of the finished area. However, my confidence diminished as I pondered the process, so I chose another stick to work with. It had its own issues, but I will not know the outcome until further along the process.
Hand plane a flat and perpendicular edge for joinery
I jointed the rear legs enough to have reference faces, then used a hand plane to create a flat and perpendicular area for the joinery. Dadoes were cut on the table saw to fit notches at the rear of S3. When a tight fit was achieved, a round over was cut at the router table. The round over and dadoes match the notches and rabbets on S3.
Dry assembly (sans back support)
Front legs were similarly created other than this joint is less secure, being only captured on two sides (rather than three on the rear legs). Cutting those dadoes and round overs only required one good corner. After test fitting, the entire stool was dry assembled for a photo op. Not so elegant at this stage, but the foundation has been laid.
I cut the bevels the opposite of my intentions and the clear chalk marks. How does that happen? Life mysteries.
S3 and outer edge of S2 and S4 receive bevels to roughly create a deep saddle. Like any prudent craftsperson, I marked the bevel orientation with highly visible chalk marks and then proceeded to ignore those markets. Cutting the bevels opposite my intended orientation. A solution was to flip the underside to the topside. Besides the grain being different than I had chosen, the new top side has a large funky area that I had determined would be a non-issue on the underside. Hopefully, it does not go deeper than my planned sculpting. This is not the first time I’ve messed up bevels. I guess bevels at the table saw befuddle my brain.
Notches for leg joinery cut at rear of S3 with dado stack
S2 and S4 also receive the bevels on the outer edges. Those went as planned. Notches for the saddle joints at the rear of S3 and forward, outer corners of S1 and S5 were cut with the dado stack. The S3 notches required holding the workpiece vertical, however the generous flat combined with thick stock made this feel secure enough for my standards. The dado stack was set up for 3/8-inch bites so I nibbled to the layout lines.
After the notches were cut I used a rabbeting bit at the router table to create shoulders matching leg dimensions. The seat segments were oriented and marked for floating tenons. Those tenons are made from 3/8 x 1 x 2 inches Baltic Birch plywood. [Note: 3/8-inch plywood is less than 3/8 inch thick, so I used 1/2-inch thick ply and reduced to 3/8 inches using the power planer.]
Mortises for floating tenons on seat segments
Mortises were cut with a 3/8-inch spiral uncut bit. A jig would be appropriate, but I chose to balance the router on the workpieces as best I could. Those on perpendicular edges were captured flush on the Moxon vise for solid support. The bevel edges were trickier. I did the first ones using the end vise and battens. Subsequent ones were placed proud in the Moxon which felt more secure and at a better work height for the task. The mortises were squared up with a chisel and the seat dry assembled with floating tenons and a rough shape was drawn with black marker.
Walnut slab broken down for easier hauling and handling
In 2013 I built a sculpted stool based on Charles Brock’s Bow Tie plans. It has received many positive comments over the years. Now I am revisiting that project as a gift for a friend. I found a large slab of walnut at Santa Barbara’s LocalWood.net. Rob Bjorklund operates this business high in the Santa Ynez foothills along the old pass road. You may wish to use the link to learn more about his operation.
I was able to find the components in the longer/narrower sub-slab. The many artifacts including extensive worm holes may cause me to change the selections as I cut into the material. We shall see. Worm holes can be small on the surface and lead to major cavities as you dig further.
Stool with lumbar support, 2013
This “simple” stool [photo on left is the first version] consists of four legs, a seat and a back rest. Easy. But the joinery is more complex than traditional mortise and tenons or dovetails. Then the fun part is sculpting. I will finally be able to use my Claire Minihan travisher that has patiently awaited its first serious mission.
I marked the components then used a jig saw for rough cuts. When possible they are well over-sized to allow adjustments. These elements were further cleaned up at the band saw.
First up is the seat consisting of five elements. One face and one edge were jointed. The center seat section (S3) was too wide for my six-inch jointer so I jointed it with some overhang. That proud material was brought to jointed face level with hand plane. A power planer let me create flat reference faces and the table saw used to rip the opposing edges. To save material I did not make the faces perfect, just clean enough to work on the joinery. There is no need for super clean surfaces with all the shaping to follow.
After this milling I restaged the seat components and identified orientations for subsequent steps.
