BENCH END VICE – Rowden Atelier


Last month we talked about making a cabinetmakers workbench. This month we are going to make and fit a cabinet maker’s end vice. The end vice is a little different from the usual carpenters front vice. This is usually metal and in our case, fitted with wooden jaws to avoid bruising the work. These are quite useful vices, but they don’t fulfil the need that a cabinet maker often has of laying a job on a perfectly flat benchtop and then securing it in such a way that one can work on it unimpeded by clamps and fixtures. The End Vice does this by a system of dogs. These are pegs that can be placed, one in the moveable end vice and one at some distance along your bench to suit the length of your job. So you bench finishes up with several holes along the benchtop. There is usually a single peg or dog that slides into these holes and is moved along the bench to suit the size of the job being worked. We’ve taken to making circular bench dogs because it’s much easier to drill a circular hole than make a square mortice in the benchtop. A nice bench dog sits in the hole and has just enough friction to allow you to push it down level with the benchtop without it falling out and clattering on the floor around your feet.

This End Vice and system of the bench dogs is just so sweet and simple and until you have used one it’s difficult to appreciate why one should go to so much trouble – but let me try and explain.

This is usually achieved by putting a spring in the side of the dog. We’ve taken to making these out of bits of coat hanger that protrude out of the side of the dog and are squashed as the dog is pressed further down into the benchtop. This End Vice and system of the bench dogs is just so sweet and simple, and until you have used one it’s difficult to appreciate why one should go to so much trouble – but let me try and explain. Firstly, imagine the problem of actually planing up a piece of wood. You maybe have machined it square or round, but you want to take out the planer marks. Now you can do that without all this kerfuffle against a stop fitted in your benchtop, but invariably you have to plane in line with the bench stop otherwise your work moves around and moves away from you. OK, you can have two parallel bench stops and put a small piece of wood across them and work your job against that, and this is how many carpenters have worked with a simple carpenters bench. Still, when the end vice is clamping your job, you are not just resting it against an end stop you are holding it at both ends securely down on to your perfectly flat bench surface. This is another thing – if your job is thin be careful when you run a plane over it because any imperfection on your bench surface will translate to the surface of your job for your job will bow into any dips and hollows in the benchtop. No matter how flat your plane is, you won’t plane a flat surface in that job.

Making an end vice seems to take someone new at this business somewhere between one and two weeks, it usually involves a couple of hours scrabbling around sitting under somebody else’s bench to work out how the damn thing fits together. Once they have done that they seem to change their sizes and proportions according to what bits of wood they have leftover from the actual bench construction. The size of the vice isn’t critical. Make it a length suitable to your requirements. Make it a width suitable to your requirements and the materials that you have available. We’ve given you a set of working drawings but use them as a basis rather than as an exact template for what you do. One of the things that you do need to copy is a massive dovetailed joint between the front member and the two pieces running across the width of the bench. If you think about it, that tail vice is subjected to enormous racking stresses. We had a student come to us with a bench already made, and he already had an end vice but had dowelled this joint together. Although the sections of timber were pretty large, the joint broke when he started to do some serious work with it.

bench end vice

Cutting this joint is a bit of fun. It’s a straight forward through dovetailed joint, but it’s on such an enormous scale that it’s hardly like any other dovetail that I’ve ever seen or made as usual make the tails first. They can be sawn out either with a decent handsaw or tenon saw, or they could be done on the bandsaw. Set the bandsaw up very carefully so that the saw is cutting at 90 degrees to the table and then buzz down your gauge line. The essential thing is that you are sawing straight. It doesn’t really matter at what angle your tail is scribed as long as you can bandsaw a straight line you can do this on the bandsaw. You can practice on a bit of scrap first that usually helps you get your eye in, and you get the feel of how the bandsaw is cutting. Next, clear your waste with a coping saw and then pare back to the line. On the fitting piece of your job mark out the pins, then set the bandsaw up to the appropriate angle. We found that the bandsaw is by far and away the most suitable piece of equipment for cutting this part of the job. You can get a very accurate result by carefully setting up the band saw table to the correct angle, then very gently buzzing straight down the gauge line. I say buzzing straight down the gauge line, but really you are cutting with the kerf of the saw on the waste side of the gauge line, almost just leaving that gauge line still there on your job. What you don’t want to do is to leave a mass of waste that you are going to have to pare away. You want these two components almost to fit together straight off the bandsaw. When you bring them together, there is a massive area to fit, and you spend a merry weekend paring away these surfaces till they go together if you don’t have the courage to go for the line. If you overcook it, however, you will be left with a rather sorrowful gap, so it’s worth learning how to use that bandsaw and buzz down the right side of the gauge line. If you do get a gap that you decide you can live with there is a way to deal with it. Once the job is assembled, and the glue is dried, take your finest dovetailed saw and run the saw down your gappy joint. What you have done is turned your uneven gap into a nice straight parallel-sided gap. Now find a piece of veneer of the same species of wood that your end vice is made from (we use Iroko a lot for these components). Try the veneer in the slot. It should be just a tad too thick to go in the kerf of the saw cut. If you take a solid flat piece of steel like the anvil on the back of an engineers vice and a hammer you can thin that piece of veneer to approximately half its thickness just by hammering over the surface. It should now easily fit into the slot. Now coat it in PVA glue on both sides and very quickly put it into the sawn kerf. What will happen is the moisture from the PVA glue will expand the veneer back to its original thickness, so you have to move quite quickly, but the veneer will find itself trapped in the saw kerf and the glue line will be almost invisible. What you will see is the different grain of the veneer because it’s almost impossible to match the grain direction exactly, but the mend will be neat and craftsmanlike. It will also help strengthen the joint. Fitting veneers into badly cut joints is not a practice I advocate, but it is something we have to do from time to time. Human beings make mistakes. A well-made mend is I think a part of good craftsmanship.

If you overcook it however you will be left with a rather sorrowful gap so its worth learning how to use that bandsaw and just buzz down the right side of the gauge line.

Once assembled, the end vice should be fitted to the underside of the bench and run into the grooves and rebates provided. Make it as snug as you can then put some candle wax on the runners. This vice can take an awful lot of abuse, but if its well made it can wear and last for a very long time. Make it badly, and the runners will allow the end vice to drop slightly as it is opened further and further from its closed position, so compromising your perfectly levelled benchtop.

We are in the middle of the run up to exhibition time at the moment. We have got two chairs approaching completion and a deadline of a week on Friday. With luck, Martin will have finished one chair by Friday giving me three days to work on the finish before it goes off to the photographer. I’m playing around with a Gesso finish which is essentially hide glue and chalk powder. However, it can be built up on the surface of the timber to create almost porcelain smooth like surface. On this smooth polished surface, I am going to play around with my watercolours and my airbrush. I don’t know if it is going to work, but this is a speculative piece, and my feeling is that I must try something new like this or there is almost no point in making this speculative piece of furniture. I feel very comfortable working with old mediums like watercolour and Gesso. Both I know are well tried and well tested, and it will be nice to use this in a contemporary furniture design. One idea I got has come from watching some rhubarb growing in my vegetable garden. The shapes and the forms of the chair legs have the same structural form, and I’m going to play around with pale green and brilliant red watercolour. If the Gesso works as well as I hope it will absorb the watercolour as it is sprayed on rather than letting it dribble and run down the chair leg like an incontinent older man. But I’ll let you know more next month if we manage to bring it off.

Nick’s also got a delicious chair to finish off. This has been one of those jobs I have been putting off for nearly two years now. I knew that there would be a good chance that I might lose money on this job and that’s probably a part of why we’ve been putting it off. But I still wanted to do it because it’s such an interesting idea; however, like all unresolved design ideas, it’s not an easy one to make. I’ll tell you more about this one too next month.

Written by our founder, David Savage. Originally published as Bench End Vice in Good Woodworking Magazine / 2001