Shims Refit the Mortise for the Neck—from the July 2001 Journal

David Brownell, Ann Arbor, MI

When resetting a loose neck, it is not unusual to find that the fit of neck butt into mortise is a sloppy fit; there actually is a gap left alongside the neck butt. In other cases, the edges of the mortise in the upper block have become damaged so that the fit is poor. Very likely, the poor fit of neck into mortise is the reason that the neck is now loose.

In the days of the Amatis and of Stradivari, the Baroque necks of violins, violas and cellos were surface-mounted in the same manner as the necks of lutes and gambas. Rather than being fitted into a mortise, the wide end of the neck was shaped to fit against the rib surface, butt-glued with hot hide glue, and three or four iron nails banged through the upper block into the neck butt. Instruments still exist which show the marks of these nails in the original block, or even the nail heads themselves. A very few are still fitted with the original Baroque neck.

Even in those days, the string tension on the violone (the three-string double bass of the time) was too high for this sort of neck installation. Instead, the neck of the double bass violone was mortised into the upper block with a sliding dovetail in not too different a manner than we do today. By the 1740s, the transition to more steeply angled necks had begun, in search of a brighter sound, and the Kloz family in Mittenwald had begun mortising even violin necks into the upper block.

Exactly fitted in this way, the join of neck to body is quite strong. Sacconi, as a demonstration, would finish the fitting of a cello neck into its mortise, and slide the neck into place dry. Then with the cello neck not yet glued, he would string it up and play on it! But if the fit of the neck is sloppy, the glue bond will fail wherever there is a gap, so that the neck comes loose. Many commercial instruments have this problem, as do those where the craftsman was simply not as skilled as he could have been.

There are three methods of adding wood so that the neck fits solidly into the mortise in the upper block of the instrument. Note that wood, not paper, should be used. Paper makes a fine temporary separator for components which are to be popped apart later. As a structural load-carrying member, a paper layer will fail every time by splitting down the middle of the paper layer. Used as a neck shim, well, the only question is how long before the paper layer will fail.

The first and most secure refit method is to clean up the edge and gluing face surfaces of the mortise in the upper block, then glue in a solid piece of new wood (either spruce, or something like willow or poplar). This can be securely clamped in both directions, so that there is a good glue bond against both edges, and against the gluing face as well. I use hot hide glue for a thin glue line and no creep. Elmer’s white glue or other PVA glues such as Titebond are not suitable for this application because of their tendency to creep under load. When dry, the mortise can be newly carved out of the added wood. The opportunity can be taken here to shift the position of the neck, its angle or alignment, or to alter the vertical tilt of the neck butt. Once carved out, very little of the new wood may remain, but what is left is now firmly glued in place virtually as part of the original block.

This is an excellent method where the original mortise is the wrong taper, or off-center. One example might be the case of replacing a clubby, oversized original neck butt during fitting of a neck graft. It obviously makes sense to shape the new neck butt to more accepted proportions, but then the oversize hole in the ribs and upper block must be made smaller. Chose unfigured willow or poplar for this added wood at the mortise. The lack of figure makes faking the varnish of the visible raw new wood to match the original rib easier; much easier than trying to match exposed spruce to the adjacent maple rib. All of these woods will glue more securely against the original spruce or willow block than will small bits of maple.

When seating the neck into the newly fitted mortise, have all of your clamps and pads selected and ready so that you can work quickly while the glue is still hot on the surfaces. You must get the neck placed and the clamps home before the hot glue cools to jelly, or the glue bond will not be strong. I use a padded caul under the button of the back, a wood wedge over the neck to create a clamping surface parallel to the pad under the button, and a Pony clamp to press the neck down into the mortise. For cello and bass, I place a nylon strap clamp threaded under the back, up over the C bout corners (pad the plate edges here with shaped wood or folded cardboard) and back across the upper bout edge to run up the upper ribs and over the toe of the neck butt. This, when tightened, snugs the neck butt against the gluing face of the upper block mortise.

Where the original mortise is in good shape and the taper matches well to that of the neck butt, a slight gap may be taken up with spruce or pine shavings. This is the second method. Shavings work best when fitted as symmetrical shims on each edge of the mortise. Choose a shaving thickness which will fill the gap up on each side so that when the neck is dry fitted with the shims into the mortise, a gap of about 1/8 inch is left between the toe of the neck butt and the button of the back (this is about twice what we would leave when fitting a violin or viola neck). Cut the shaving to a length just a bit less than the edge of the mortise, it will tend to slip down a bit as the neck is seated and so needs about 1.5 to 2mm of room at the bottom. The two things to avoid are trapping a fold of the shaving shim between neck face and that of the mortise, and trapping the end of the shaving under the neck toe. Either will prevent the neck from seating properly. While it is possible to use shorter bits of shaving to tighten the fit at the toe end of the mortise (the spruce will compress to adapt to slight imperfections of fit), any deviation more than the thickness of a file card is best filled in using solid wood as the shim. Then the mortise can be re-cut to exactly fit the neck heel.

Arrange your clamp pad, wedge for over the fingerboard, and suitable clamp(s) alongside the workbench. Wet the mortise edges with hot glue, then the shaving, and place the shaving onto the mortise edges so a gap of about 1/16 inch or a bit more is at the bottom of the mortise edge. I use the glue brush and a hot palette knife to maneuver the shaving. Wet the face of the neck butt and the matching face of the mortise with hot glue, then brush more hot glue back over the shavings and the edges of the neck butt. Slide the neck into place, verify alignment, and clamp. Once aligned and clamped, you can then score the excess shaving exactly at the rib surface with a sharp knife, and lift the surplus away from the sides of the neck heel. This comes away more easily before the glue dries.  As a variation of this method for a somewhat larger gap, willow or poplar strips can be glued into the mortise in place of the shavings. The procedure differs in that the strip shims are glued into place by themselves first, using prop sticks and wedges cut from scrap wood to press the shim wood firmly against the mortise edge at top, bottom, and middle. We want no gaps. When dry, the edges of the shim are shaved to match with the rib surface using thumb plane, chisel and/or knife, and scraper. Pare out the mortise to make a firm fit to the neck, the dry gap between neck toe and back button should be about the same as described above. Small adjustments of neck alignment can be made using this method, which are not practical when using shavings. If larger shifts of alignment are needed, then a solid wood fill of the mortise is better.

The third method involves the inlet repair of damaged edges or cheek faces of the mortise. While loosened splinters can be glued back into place, a thin shim over any gap is literally just ‘papering over’ a weak spot. A chewed up or badly splintered edge can be pared out of the upper block using a long mortise chisel, or the edge of a broad chisel working down from the rib edge. Clean the remaining block edge up to a solid smooth surface. The important thing is to get clean, smooth and solid surfaces to which you can get a secure glue bond. Sometimes it is sufficient to use a chisel to pare out a clean bevel space in the block wood under the rib end. In other cases you will need to chisel out parallel with the grain of the block and perpendicular to the mass of the block. A wedge-shaped insert can be exactly fitted to a bevel space under the rib edge and against the front face of the mortise. If the bevel angle is less than 30, this will glue quite securely. If the angle is undercut as much as 45, then pressure from the neck may tend to lift the rib. Shave the block wood cleanly from the inside surface of the rib end if necessary.

New matching wood is now exactly fitted to the space pared out of the mortise edge, leaving excess wood into the final mortise space. Try to get a precise fit of the added wood against three surfaces: the new edge which you have just pared smooth, the front face of the mortise (you should have carefully extended the mortise face in the same plane as you pared away damaged wood), and against the underside gluing surface of the rib (the rib face that had been glued to the original block). You likely may need to do this on both edges of the mortise. Match the grain direction of the original block wood with your new insert. After gluing the added pieces with hot hide glue, then the mortise is re-cut to exactly fit as described with the other methods. While the glue bond by itself is quite sufficient for cello and smaller instruments, the high string tension of a double bass often makes it necessary to additionally secure this added wood to the original block with screws.

BACK TO VMAAI JOURNAL