Rebuilding the Æolian Orchestrelle
58-Note Player Reed Organ, 1912 Model "W"
by Richard Vance
Copyright (C) 2002 by Richard Z. Vance
rev. A, 3 October 2002

Chapter 21.5 The Pallet Tiewires

    Now is the time to connect the wires from the pouch followers to the leather padded staples on the pallets.  It is quite critical to make these connections the correct length; too short, and the pallet will not seat firmly, producing a cipher.  If the wire is too long, the pouch follower will bottom before the pallet is open far enough.

    Step one is to turn the big screw eyes near the heel of the pallets crosswise, if that has not already been done.  These eyes only prevent the pallet from bouncing up off the rear pins during rapid play.  They do not function as fulcrums, and should not be tight to the pallet heels.  If there is the slightest contact between the eyes and the pallet sides, they will "cock up" the pallet slightly, preventing full valve seating at the top, and causing a hard to find cipher.  Look carefully at every eye, and make sure there is 1/16" clearance between the eye and the edge of the pallet.  Unscrew the eyes a half turn, or bend the eye laterally, if necessary.

    Next, make sure all the pallet springs are in the pallet grooves.  One tries to avoid bending the springs when they are turned aside to remove the pallets, but some will inevitably be bent a bit, and have the wrong tension.

    In order to check that the springs have the correct tension, I made a spring force tester, using some strips of hobby brass and a small coil spring.

    This is used to measure the force needed to open the pallet against its spring.  The actual magnitude of this force is unknown; I had no way of measuring that.  So to set the spring in the tester, I tried it on a number of pallets, assuming that many springs would still be correct, on average.  The tester spring was set to pull the pallet about 1/18" open, on the "good" springs.

    Al the pallets were checked.  Where the tester pulled the pallet to far open, the pallet spring was bent forward (up) near the place where it was driven into the wood, using the little slotted tool from below.  Another check is to pull the pallet open and let it snap closed (without the wire connected).  The pallet should snap closed with "thwap" sound that one soon learns to recognize.  This sound indicates that the pallet is closing firmly, all around its edge. A softer noise, or the observation that the pallet is not moving closed perfectly parallel to the plane of the valve board, indicates that the spring is not pushing in the very center of the pallet, at right angles to the pallet's movement.  This is fixed by bending the loop of the spring sideways, left or right.  Sometimes the spring loop appears to be not quite centered on the pallet slot when the valve closes properly, but this is ok.  This must be due to some slight sideways curvature of the long end of the spring.

    Another check is to open the valve and look at the surface of the valve leather with a strong light.  The oblong outline of the valve hole should be firmly imprinted on the leather surface, evenly all around its edge  Now one can correct any slightly bent pallet guide pins by observing if the impression of the seat is more or less centered on the leather. Out of the 348 pallets, I found two that would not pass this test, no matter what I did.  Upon removing these pallets, I discovered that there was a slightly hollow soft spot in the felt layer; a defect that had escaped my notice when releathering the pallets.  I had to replace these seats.

    The first step of tying the wires is to pull the wires straight from the pouch follower staples, and rub out any kinks on the wire.  Make sure that the leather tab on the pallet staple is still inside its wire loop.  Sometimes the tab had slipped out of its loop, and had to be pulled back inside with al long tweezers.  Cut the wire to a convenient length, about one inch beyond the edge of the valve board.

    Make sure the leather tab on the pallet staple is turned towards you, centered on the staple. Feed the wire up through the staple, from below.  Hold the pallet tightly against the valve board, and pull the wire tight.  Pull the free end of the wire under, across, and back up over, to form the first turn of the loop.  The wire loop should be small enough to capture the small center of the tab fairly tightly, so the tab can not slip out sideways.  Now one can see why Aeolian took the trouble to laboriously install these tabs.  The pouch leather is thick and flexible enough keep the loop loose enough to rotate freely around the staple, at the same time eliminating any lengthwise slop on the wire.

    At this point, the wire is "too tight", since the initial pull-through of the wire has drawn the pouch all the way out.  Now one establishes the initial correct length of the wire.  Holding the pallet against the board, pull center of the the wire sideways, about 1/4". This will unwind the loop a small amount, leaving some slack. Wrap the free end of the wire three more times, to complete the tie.

    Now for the final functional check.  The pallet should now rest flat against its seat.  Retest the tension of the spring with the tester.  The pouch should now not appear tightly pulled out, but should be not quite all the way out, showing a slight depression around its edge.  Push the pouch in with a stick, and let go.  The pallet should close with the same thwack as before the wire was attached.  The valve has to open only about 3/16" for full air flow to the reed.  Sometimes it will be necessary to lengthen the wire a bit more by pushing it sideways again.  Occasionally some binding of the wire loop round the pallet staple will be detected, if the first loop was done too tightly.  Poking the loop gently with the point of a bamboo skewer will fix this.

    Since photos don't show the procedure all that clearly, I have replicated the steps in the drawing below:

    This procedure seems inordinately tedious, but it actually goes very fast, once one gets the hang of it.  Here is another example of Aeolian's cleverness, enabling the low paid workers at Meriden to get good results without special skills.  New, accurately machine made springs; the leather tabs; and the use of waxed, cotton covered pipe organ wire that ties firmly but unwinds easily when pulled sideways, all contribute to reliable, repetitive work.  I suspect that tension testing of the springs was not required.  A good valve has a feel and sound that is unmistakable, once one has done a few hundred of them, and any fault can be easily corrected.

[back] [forward] [index]