Orchestrelle co. 65/58 note push-up player

This sequence of pictures show the piano player in the state in which it arrived. The case was crazed and scratched with pieces of veneer missing and (after over 100 years) everything was filthy. The hard felt "fingers" on on the ends of the bell cranks were missing, and the bell cranks themselves had come unglued at the angles. The motor looks in reasonable condition apart from the cloth but had started to disintegrate. And of course the pneumatics and tubing had completely perished. But all those were the least of our concerns as we found out later...

As I was very out of practice with pneumatic work I started the restoration with the most self-contained unit to do in isolation, namely the air motor. The air motor looked in reasonable condition but quite literally fell apart on removal from the player. The pneumatic boards however remained firmly attached to one of the skins of the motor body. Hopefully my pneumatic work is now neater than it was on these, the first pneumatics I'd re-covered for many years.

The main chest (containing the primary and secondary valves which respectively detect the perforations and operate the striker pneumatics) gave little indication of what was actually inside until the front was removed. It was necessary to dismantle this as everything was leaking extremely badly and most of the valve facings, all the gaskets and some of the pouches needed replacement. The structure is actually a ladder-like assembly of pouch boards and valve boards (three pairs, one above the other) with the main wind connections between the different layers at the ends. This proved extremely difficult to dismantle for repair, and it was necessary to split one of the valve boards away from the back using various thin knife blades. Some damage was done but with care the split wood could be reattached to its parent material. the final image shows the reassembled chest with new gaskets.

The pneumatics were glued directly to the stack boards and proved difficult to remove by the textbook method of prising them off (many were damaged and some needed a new board to be made). It would have been better to remove them using an iron to soften the glue; this proved far more effective on the last few pneumatics.

The pneumatics were originally covered in a white rubbererised material, presumably jaconet, but the rubber coating had completely disintegrated and presented no airtightness whatsoever. Some of the pneumatics had tan pouch leather at the hinge end instead of jaconet. Some of them were better made than others - one of them was highly asymmetric! All the pneumatics had a hinge spring, but these were in poor condition so new ones were made from piano wire. The pneumatics were recovered using conventional rubber cloth and hot hide glue, and glued back onto the boards with a gasket of hard paper. Care was taken during reattachment to avoid creasing the pneumatic cloth - padding can be seen being used to prevent the pneumatic being squeezed fully shut by the spring clamps.

The valves and some of the pouches were in poor shape. The primary valves themselves simply needed refacing (and some valve stems replacing) and are not shown here. The primary pouches seemed weak, several were split and were all replaced using ready-punched discs of tan pouch leather.

The secondary valve seat facings were one of the main reasons for the overall failure of the system and accessing them for replacement was difficult (see the article in the Journal). Replacing the valve seat facings required some ingenuity, with the new facings being held on weakly adhesive "sticky note" backings during manufacture and gluing to prevent them from distorting. These were pressed into the valve wells using a heated bar, and after the glue had set the backings were peeled off. The secondary valve stems were badly corroded but it was impractical to dismantle these and replace the upper valve faces, so the corroded stems were cleaned up and the upper facings retained as being just usable.

The secondary pouches were mostly just about acceptable and I only replaced the few pouches which were damaged, though in retrospect I probably should have replaced them all. The edges of some of the pouch wells were very thin indeed, due to the pouch board being planed a bit too much when the system was assembled originally.

The finished player works well and is an effective if very primitive instrument, with none of the usual features to aid in expression. All dynamics must be applied by variable pumping of the pedals, and all accenting can only be applied by "jabbing" the pedals. There is no split stack to enable treble and bass to operate at different steady dynamic levels, and there is no automated theming system. There is a primitive manual lever for accenting the whole compass but it is of little practical benefit and is not currently operational.

The sustain ("loud") pedal shoe was missing, so a replacement was made out of mild steel. Also, the rear panel was too badly damaged to be useable (it seemed to have been used as some kind of rolling board and was cracked across the grain of the veneer) so a new rear panel was made from plywood, stained and varnished to approximately the correct colour, and the original panel used as a source of veneer for patching the body of the instrument.

The decision was taken to conserve the original polish and revive it as far as possible, where necessary cutting it back with fine abrasive paper and re-polishing with button polish (a form of French polish). This was partly because amateur attempts at stripping and repolishing rarely give entirely satisfactgory results and destroy the originality, and professional re-polishing would have cost more than the value of the instrument would have merited. This policy has been largely successful though the family pet has been diligent in one place in "distressing" the veneer and polish repairs! (This has now been repaired and she's not repeated that vandalism). Some further improvement has been made to the finish since these photos were taken.