Assembling the jacks

So far, I've assembled a couple of prototype jacks in testing out the jack design. Now that the process is being done for real with actual jacks, here's the procedure.

First, any burrs in the jack body caused by drilling the axle hole are trimmed with a small chisel:


Next, the axle hole in the tongue is enlarged with a miniature reamer:


The looseness of the reamed hole is checked with an axle pin held in a pin vise. The tongue should hang down freely under its own weight:


The jack is temporarily assembled to check the fit of the tongue within its slot:


If necessary, the tongue is sanded a little to narrow it for a looser fit:


Once again, the tongue should hang down freely under its own weight:


3 mm strips are cut from a sheet of 0.005" brass shim stock to make leaf springs:


Individual 19 mm long springs are cut from the strip and slid into the spring slot:


The 3 mm width was chosen to make the spring fit snugly within the slot.

Each spring is pre-tensioned by bending it forward:


Finally, the jack is reassembled:



Each assembled jack is tested to make sure the spring returns the tongue fully to the forward position. At present the springs are stronger than necessary, but they can be adjusted later by bending them back slightly.

Drilling for the axle

The tongue is held in its place within the jack slot by a little axle, which acts as the pivot point for the tongue.

Here's the setup for drilling the axle hole:


The axle hole is made by a #61 drill bit. A tongue is placed within the jack slot, then both are laid face-down on the board and slid to the right until the tongue touches the little nail shaft. At that moment the tongue has also bottomed out in the jack slot. To make sure the tongue is centred within the thickness of the jack, a few slips of index card fit inside the jack slot and shim the tongue upwards. Then the jack and tongue are drilled simultaneously all the way through.

Next, the tongue is removed and its hole is enlarged with a little reamer to enlarge it slightly. This ensures the tongue pivots freely on the axle.

Italian jacks often use a flat brass spring to push the tongue forward once the plectrum slips under the string. The bottom end of this spring rests in a slot of some kind. On some jacks I've seen, a groove was cut in the back of the jack and a slightly thinner piece of wood glued in, leaving behind a little slit. I thought a simpler approach would be to use a diamond burr to route a tiny dovetail slot down through the jack slot for a distance of about 10 mm. Since the slot is narrower in front, the spring can't fall out. A piece of 0.005" brass shim stock is cut to fit this slot, and is made long enough to come up to just below the plectrum mortise. The discrepancy between the bevel angles of the tongue base and the bottom of the slot in the jack gives a little extra clearance to the spring as it comes up behind the tongue from the dovetail slot.

Finally, the tongue is replaced in the jack slot and a 1.0 mm nickel-plated axle pin is pushed through the jack and tongue. The pin is a little shorter than the jack width so there is no excess to trim when it's pushed home.

The business end of a test jack, using a scrap tongue that is punched and holds a plectrum:


Finding the correct position of the axle is critically important. A low axle lets the tongue tilt back quite easily: too far back, in many cases. Then the tongue bounces off the string behind it, creating unwanted noise. There is also the problem of a low axle allowing the tongue to kick backwards just as the plectrum plucks the string. This represents wasted energy that should have gone into the string. In such a scenario, the exact moment the string is released is governed—wrongly—by the jack design, not by the characteristics of the string and plectrum. Worst of all is the lack of fast repetition of notes when the jack has an excessively low axle position.

A higher axle is the way to go, even though it makes the tongue less willing to tilt back. In this case, the spring must be carefully adjusted by experimenting with the spring strength and the position at which it contacts the back of the tongue.

I put together a few trial jack tongues and bodies and drilled the axle in several locations, starting with a position twice as far down as the plectrum mortise (19 mm, in other words). Then I put each trial jack into the harpsichord, installed a plectrum and played notes in various parts of the compass, observing the tongue behaviour. After moving the axle up slightly with each new attempt, I found a position at which the tongue kickback stopped. I also found I had to drill the axle closer to the front face of the jack: