Saturday, April 23, 2011

Building A Bench On Bench

In Roubo's Slippers, I fine-tuned the height of my workbench for planing. However, many other operations benefit from a higher work surface. Moxon's double-screw vise is one way to elevate and hold pieces for dovetailing.

I first stumbled onto the idea of a benchtop mini-bench last year on Alf's blog. She had built one based on Jeff Miller's 2005 Fine Woodworking article. Miller used veneer press screws to form a double-screw vise on the front. Alf used wooden screws she turned on the lathe and threaded with a threadbox. Either way, this adds the elevated work surface to go along with the vise.

Then I saw this one on Joel Moskowitz' blog at Tools for Working Wood. It's built by the Philadelphia Furniture Workshop, their version of Miller's design. You can buy either a finished one or the hardware kit (screws and clamp handles) to build one from Joel here; instructions for building a simple plywood base are here.

After debating back and forth which one to build, I decided on the last one. Miller's was a simple slab design, but I liked the PFW butcher block construction, as well as their base. I used standard veneer press screws, as Joel showed in the initial PFW version. Links for suppliers are listed at the bottom of this post.

The one technical challenge was figuring out how to get the long veneer press screws into the laminated top using hand tools. I didn't think I'd be able to hand-drill parallel 9"-deep holes straight enough. I thought about chiseling channels into the underside. I initially expected to fit the flanges for the screws to the front. Then it hit me: drill the holes one lamination at a time, and bury  the flanges in the first lamination.


Marking out the first rip for the butcher block laminations in a slab of 2x10 poplar.


Ripping the first piece. Since this board is flat-sawn, when I roll the pieces 90 degrees they'll be in quartersawn orientation for gluing up the top. The kneepad makes this much more comfortable.

Make sure your saw is sharp for this kind of work. Do yourself a favor and spend 10 minutes sharpening it; you'll save 30 minutes of heavy work.

In thicker stock like this, flip the piece frequently to keep the saw on track (mark both sides with the gauge). You can hold to within 1/32" of your line for the whole rip that way. Otherwise you're likely to find the back side of the cut has wandered by 1/8" or more, requiring a lot of work with the plane when you joint the edge for the next piece, and flatten the final glue-up.


Jointing the edge for the next rip. Notice the long-sleeve shirt is off, and the bandana is on. This is heavy ripping.


Cutting the pieces in half.


All the rough-cut pieces laid out to plan screw placement. Just barely visible, in addition to a cabitnetmaker's triangle, I've marked arrows on each piece showing which way the grain goes. They all need to be oriented the same way for planing once they've been glued up.


Carefully drilling the holes in the first piece. Each piece will serve as the guide for the next. For this piece, which will be the loose vise face, I used an 11/16" bit; the screws are 5/8" in diameter, so this allows a tiny bit of play. For the remaining pieces, I used a 12/16" bit. The extra tolerance will be hidden in the lamination, and allows for any error in drilling. The holes for the flange are 1", which required a large-sweep brace for extra power.


Chiseling out a mortise for the flange, like a hinge mortise. Notice the side ridges down its length; I ran a gouge down each side of the hole for them, then flipped the flange over and dropped it into the hole. Fitted into the mortise, the flange face was flush with the surface of the wood, so it wouldn't interfere with the glue-up.


Loose test fit. The rightmost piece is the vise face. You can see the flange buried in the second piece. The left piece will slide up the screw to be glued to the second, and so on for the remaining pieces.


With the screws threaded into the flanges and cranked in, checking the alignment of all the drilled pieces (the screws don't reach the back four pieces). Nice smooth action, no binding anywhere.

At this point, I removed the screws and glued up the assembly. I replaced the screws and adjusted the pieces in the clamps to make sure they continued to turn freely.

While this dried, I started on the base, using some scraps of Baltic birch plywood. It consists of two I-beams joined by a crosspiece. Each I-beam is made of three pieces, top and bottom plates with center upright. The center fits in dadoes down the length of the plates. The height of the center determines the height of the base.


Cutting the I-beam plates.


Scoring down the length of a plate for the dado.


Sawing out the dado sides. Because this is plywood, I switched between rip and crosscut saws for each ply. That kept the saws from catching in the alternating grains.


Bevel down, chiseling up chips for the length of the dado.


Still bevel down, running the chisel down the dado to clear the chips.


Now with bevel up and chisel flat, paring the dado floor smooth.


When I test fit a piece in the dado, it was too tight, so I widened it ever so slightly with a #79 side rabbet plane.


Test fitting a piece into the widened dado. A good snug fit.

Once the top glue-up had dried, it was time to flatten it. First I flattened the bottom side with a #5 jack with cambered blade, followed by a #6 fore plane. Then I did the top side.


Diagonally with the #5 across the rough surface of the laminations to take it down quickly and evenly. This produces rough chips.


After some diagonal passes with the #6, straight down the length. Once I was getting consistent full-length, full-width shavings, I was done.


Using a scraper to deal with some nasty reversing grain that tore out in one corner.


Trimming the end even.


Shooting the end grain.


Chamering all the edges.


Drilling dog holes in the top.


Testing for height. This will determine how I size the I-beam center pieces for the base. With a Gramercy holdfast dropped all the way down and clearing the bench top, this height just has my elbows at 90 degrees. That's a recommended height for carving, and should be fine for dovetailing.


Final assembly of an I-beam. Modern quick-change combination bits work fine in an old brace.


Installing screws with a Yankee driver. The white block is wax for the screw threads.


Setting the screw with the quick-change bit reversed to its driver head. Just be careful, because this produces enough torque to drive the screw right through, damaging the work.

Rather than attaching the base permanently, I installed threaded inserts into the bottom side of the top, and screwed knurled knobs in through holes in the base. Getting all the inserts to be in alignment is another challenge. I ended up having to enlarge a couple holes in the base with a chainsaw file so that everything went together well. But this allows me to build a shorter base to swap out with this one should I find it necessary.

The last step was to make leather washers to go up between the screw handles and the front vise face. Otherwise, with the screws tightened, the handles rub against the face. I initially made wooden bushings, but one cracked when I tightened the screw down on it.


Punching out a leather ring with a gouge.


The completed mini-bench held in place with more holdfasts, with a piece in the vise ready for dovetailing. I even went to the trouble of putting a couple coats of lovely Aged Olive color paint on the base (no, that's not the much-reviled 70's era Avocado color!). The top really looks nice. The photo just doesn't do justice to the quartersawn grain surface.

I tugged on the workpiece to make sure it wouldn't move while working. Everything held firmly in place. And when I say firmly, I mean rock-solid, like it was physically a part of the main bench. This should be a great back-saver, so I won't have to lean over my work so much.

Veneer press screws are available from
Rockler and Woodcraft. Threaded inserts and knurled knobs are also available from Rockler. Holdfasts are available from Tools for Working Wood; get 2 pairs, one to hold the mini-bench down, and one to hold the work down. (When you get them, wipe off any machine oil residue, then wrap sandpaper around the shaft and give them a spin to roughen them a bit, giving them a super grip in your dog holes.)

Friday, April 15, 2011

Roubo's Slippers

When I built my Roubo workbench, I obsessed over how high to make it. Workbench height is always a compromise, because different tools and operations require different posture.

Planing is what ultimately limits the height for a general-purpose bench. It needs to be low enough to plane comfortably, or your arms wear out quickly. However, too low and you get a sore back from stooping over too much. And a good planing height is a little low for everything else.

Because everybody's built a little differently, there's no standard measured height. Instead, it's better measured in relation to your own body. Chris Schwarz likes his bench to come up to the point where his pinky  joins his hand. Roy Underhill likes it to come up just under his knuckles when he makes fists. And Mike Dunbar likes it to come up under his palm when he tips his hand up flat.

At some point you just have to pick a height and go for it. You can hedge your bets by erring on one side or the other. Keep it on the high side with the expectation that you might need to trim a bit off the ends of the legs to lower it. Keep it on the low side with the expectation that you might need to add some kind of feet or risers to raise it. I chose the latter, using the pinky measure.

After using the bench for about 6 months now, I decided it was just a bit too low. After more obsessing, I figured raising it by two 3/4" thicknesses would be better. But I wasn't totally sure. So I came up with these slip-on feet that can have loose shims added or removed for fine tuning.

I opted for the simplest possible construction here, since there's no need for any structural strength. The feet just need to keep the shims in place under the legs if the bench slides around a bit. They're simple nailed boxes, about as sophisticated as a kid's birdhouse project.


Ripping a scrap of 3/4" CDX plywood for the bottoms. A good sharp crosscut saw cuts plywood and MDF just fine.


Crosscutting the bottom to final size. Wait, is this crosscut or rip? That's not really even meaningful with plywood.


The bottom pieces slipped under the legs for testing. Time to do some obsessing and see if this is a good height.


After cutting a couple of short sides from a scrap of pine 1x4, marking out for the overlapping long sides.


Crosscutting the sides. I didn't bother shooting the endgrain, no need for that kind of precision here.


Using a bradawl to make pilot holes for the nails. Otherwise this pine would split, nailing so close to the edge.


Driving the nails on a short side.


Nailing a long side.


An alternative for making the pilot holes: Yankee drill. Using the awl can give you sore wrist, forearm, or elbow from the repetitive twisting motion.


My concrete basement floor isn't perfectly flat, making the bench wobble (it couldn't possibly be that my bench is less than perfect). In addition to the second layer of CDX inside this foot, dropping in a layer of fiberboard to shim it up.


Levering up the bench while slipping the feet in place.


Still just a tiny bit of wobble, rectified with a few playing cards slipped into the opposite corner foot. Note that the front feet are open on the front side. On the left front foot, that keeps it from interfering with the leg vise. On the right, that keeps the plane of the bench front flat. If these feet shift around, I can put a screw in through the side.


The feet in place, all shimmed up and rock solid.

Ok, so they're not as graceful as Cinderella's glass slippers. More like big ugly boondockers. But they get the job done. I'll use the bench like this for a while and see how I like it. So far it does feel more comfortable while planing.

The final height? It comes up to the point where my pinky joins my hand. Wait, you ask, wasn't that where it was before I added the feet? Yeah. It turns out you shouldn't slouch when you're measuring for bench height.

Sunday, April 10, 2011

SAPFM New England Chapter April Meeting


Paul Lelito's Wood-Mizer portable bandsaw mill, looking out across the cranberry bog. A  holly log sits on it ready to be cut.

I got my first sunburn of the year on a gorgeous spring day at the April meeting of the Society of American Period Furniture Makers, New England chapter. The topic was custom sawmilling, presented by Paul Lelito and Freddy Roman. The location was Paul's site at the Magnolia Cranberry Bog, Marion, MA. The attendees included SAPFM members from Massachusetts, New Hampshire, Connecticut, and Rhode Island, a mix of hobbyists and professionals.

Freddy is the coordinator for the New England chapter. It also turns out he lives half a mile from me! So I offered to drive the 90 minutes to the meeting and haul lumber back in my van.

This was the rain date, since the previous Saturday had been wet snow, and Freddy needed more people to sign up to cover expenses. Quite a change in the weather. Kudos to Freddy for pulling that off!

Freddy represents the future of high-end woodworking. Not yet 30, he graduated from Phil Lowe's Furniture Institute of Massachusetts program 5 years ago. He specializes in furniture restoration and reproductions, particularly veneer, inlay, and banding, as well as commission work. If you have some fine antiques you'd like to have restored, you can reach him at 860-670-2584.

Paul graduated from Phil's program 3 years ago, during which time Freddy was working for Phil teaching and doing museum restoration. If you're looking for fine lumber, including beautiful full flitch sets, Paul has both seasoned and green, all air-dried. You can reach him at 508-451-9999. He's also available for custom on-site milling at a day rate.

Paul was offering some good deals on lumber, so the first order of the day was prowling the stacks. In addition to flatsawn flitches, he does a lot of quartersawing. That's takes more time and produces more waste, but the resultant lumber is spectacular, especially in species with distinct medullary rays.


Attendees examing some full flitch sets.

My treasure find of the day was in the covered racks: quartersawn beech, cut in 2006. I've been looking for this for over a year now to make side-escapement molding planes. I have Larry Williams' planemaking and sharpening DVDs, and a stack of iron blanks from Lie-Nielsen.


What have we here? Quartersawn beech in the center rack, well seasoned and ready to work! I managed to control myself and just buy one board.


Big gorgeous blocks of basswood. Freddy bought one of the big ones, and I bought one of the small ones. It has lovely figure in it, almost like bird's eye maple.


A large walnut log squared up into a cant. Sawing this would be the afternoon project. Paul uses a Bobcat to move these things around.

After some brisk commerce, Paul went over some of the problems and hazards of custom sawmilling. These included drying problems, items embedded in the wood, and operator errors.


Paul presents sawmilling problems to the group. Behind him is a board that twisted up like a perfect airplane propeller when it dried.


Some of the metal embedded in the walnut log: a pulley for a laundry line, and a large screw hook. These would destroy a blade. Excising them takes a big chunk out of the log, especially when the tree has encapsulated them over the years.


Paul shows the result of trying to rush a cut: a wavy surface as the blade bowed by trying to force it through. He flipped it over to show the back, where he slowed down for the next cut. That had a fine flat surface.

Next was a short walk in the woods beside the bog to talk about tree identification. Paul pointed out holly (the biggest holly I had ever seen before was just a decorative shrub), various pines, oaks, tupelo, and beech. He said tupelo was used in World War II era submarines for propeller shaft bearings. It's so tough, he said, you can bury a wedge in it and it won't crack. However, cut thinner and dried out, it eventually becomes very brittle.


Paul talks about differentiating poison sumac from non-poisonous varieties based on its buds. Good to know before you go stomping through the brush with your chainsaw!


The view back across the bog to the mill area.


The group heading back. This is such a beautiful area, and most people just blow on by driving out to Cape Cod.


Time to fire up the mill!

For the morning milling, Paul had a short but large diameter log of holly. One of Freddy's goals for the day was to get some of this to use for inlay stringing. Holly is prized for its light color. It's also very expensive.


After removing a bark slice, Paul discusses the blue discoloration that occurs rapidly in holly due to fungus. It needs to be cut thin and dried immediately.


Paul, left, and Freddy, right, taking the final slice.


Paul stood a bunch of the slices on their sides and cut them down the center to give everybody a piece to take home.

For lunch, Paul had a big firepit going with a huge pile of chicken on the grill. There were also salads, chips, cheeses, cookies, and brownies. Nobody went hungry this meeting! This was followed by a raffle, with more than enough prizes for everyone.


Jae (John A. Ewart), who is both a professional woodworker and a professional chef, tending the chicken. Curiously, the old Coca Cola sign in the background is in French.


The raffle prizes? Everyone got to pick a stack of lumber off the side of the shed. Don't let the weathered look fool you, this was all well-seasoned wood in a variety of species. A little surface cleanup will make is shine. Here Paul Williard examines the stacks. I got the stack with the short piece of walnut on the right, as well as some more beech.

After doing the first half of the raffle tickets, Paul loaded the walnut cant onto the mill and started flatsawing. The length just fit on the rails.


Taking the first slice. This model of mill has hydraulics for loading and manipulating the log, powered saw lift for precise thickness selection, and a chain-driven carriage. The kerf is wafer-thin for minimal waste. You want that in 4/4, 8/4, 12/4? No problem, name your slice.


Even though Paul had removed all the nails he could find, there were still some deeper in the log. Here he points one out to be marked in the first board. The iron in the nails reacts with the tannin in the wood to form black discoloration, almost like ebonizing dye.


Changing out a saw blade. The sawyer has to decide if the value of the lumber than can be harvested outweighs the cost of dulled or destroyed blades. Paul buys them by the box and sends them for several resharpenings, but the expense adds up. In this case, the lumber is pretty valuable, so it's worth continuing.


Paul digs out the remainder of the nail in the log as Freddy looks on. While this damages the wood, the surrounding discoloration means some defects will need to be cut out anyway when the board is cut up. When the blade hits a nail, in addition to damaging the teeth, it vibrates up and down, resulting in a rougher surface that will require more planing later.


The color on the underside of the freshly cut board has a greenish cast along with the rich brown we normally associate with walnut. The seasoned walnut in the shed had such a reddish cast that I thought it was cherry. Freddy said that's the result of exposure to light and air during drying, but it's just surface color; planing will reveal its normal color.


A dulled blade. The teeth on these are pretty sizable; I left the blower in the background so you can get some idea of the scale.

The next blade only lasted one cut, due to more nails, of larger size. Ouch! That one was too heavily damaged, completely missing the tips of at least 3 teeth and several more severely beaten, so Paul gave it to me for blade stock, along with some old already broken up blades.

Freddy and I are going to make framesaws out of it. I'll do one of Joshua Clark's Hyperkitten design, and maybe another using more traditional buckles, and see how it works out. Later I gave Paul $20 to help cover the cost of the blade because I felt bad about throwing it away so quickly.


The boards starting to pile up on the sawhorses, stickered for air circulation.


The final stack while leaving one big slab. There are 7 boards, each 28 square feet of surface. At 4/4, that's 28 board feet; at $10/bf, each board is worth $280, for a total of $1960 for the stack. I don't know what it cost to obtain the log, but it's worth the price of a few blades, even after cutting out damaged areas.


Someone wanted to buy some large square stock, so Paul stood the remainder on its side using the hydraulics and took a cut. Chef Jae took the last piece for a workbench slab top (after a few years of drying), where the nail discoloration will add character.

Here's a video of the day's work:



After this, Paul finished up the raffle and everyone left with big smiles. Freddy and I loaded up our swag, then helped Paul as he used the Bobcat to consolidate some stacks to make room for more incoming logs and get things covered up.

On the way home, Freddy and I stopped at his shop to unload his lumber. He's a partner with Will Neptune. The shop is located in Littleton, MA, right across the street from the building I worked in 15 years ago for Digital Equipment Corporation (which became Compaq, then HP, now IBM).

Freddy showed me some of his bandings. These are complex sandwiches of contrasting veneers in a variety of patterns that are then sliced across in veneer-thickness bands and inlaid into larger pieces. Making them is very exacting work. Not many people would have the patience for it.

The other thing he makes is harps! He showed me a finished one and several in progress. The work is more like boat-building than furniture: a few straight lines, but curves and odd compound angles everywhere. Placement of all the holes for string pegs and tensioners requires a large template.

Freddy really impressed me. He's very committed to his craft, and holds himself to high standards. He told me some horror stories of poor restorations he's seen that reminded me of my 10-foot dovetails: looks OK from 10 feet or more away, but don't look any closer. Freddy's standard is more like a 1-inch distance: get right up close, and you can't see where the original leaves off and the restoration picks up.

His willingness to run the SAPFM New England chapter is an example of his dedication. He works crazy hours as a self-employed businessman, yet is still willing to put in the time to make the NE chapter a success. The high-end work he does has a limited clientele in the best of times, and the current economy makes it even more challenging. But his priority for meetings is that everyone learns something and enjoys themselves.

Next month's meeting features Peter Follansbee demonstrating 17th-century rived, carved, and pegged oak furniture at Phil Lowe's shop. I'm looking forward to it! Once again, I feel so fortunate to live in an area with so many excellent woodworkers, willing to share their talents, knowledge, and resources.