I've just returned from a visit to Germany and am catching up with
MMD items about the Pistonola. As Julian Dyer said, he asked me to
do something on this mechanism for the Player Piano Group Bulletin some
time ago. I'm putting an article together, but I feel, sitting here
with a piston module in my hands, that I must jump in and offer a few
observations to be going on with. There are two key factors to
understanding the mechanism's unusual functions.
First, the primary valve is not the pin tip, but the brass cone
immediately under the bar connecting it to the primary piston. This
has been missed by several experts I've met who, hoping to 'get things
working', have consequently suggested doing unforgivable things to the
bottom pin guide which is in fact not in the signal path. The valve
and alloy seat/nipple must be very carefully cleaned of any dirt or
corrosion with a suitable solvent to enable the conical valve to seal.
Once this is done they will self-clean.
Second, and uniquely I believe, the system reads the roll at a
constant, relatively low vacuum *which keeps drag on the paper and
thus wind motor size and consumption to a minimum* but plays at
very high vacuum levels (cf. transistor emitter-follower topologies).
The low reading vacuum is initiated, established and replenished as
necessary by the primary regulator found in the upper right cheek of
the piano. This is a sensitive diaphragm-operated slide valve which
closes when the set vacuum level is attained. (In the MMD pictures
this is incorrectly labelled the wind motor regulator, which actually
is located elsewhere.)
But here's the impressively clever bit: the low vacuum is predominantly
supplied through any open primary valves by the high(er) vacuum under
the secondary valve pistons. And it is actually low primary vacuum
that lifts the secondary valve pistons which have high vacuum above
them. The loose fits of the primary and secondary pistons in their
bores provide the two essential bias (bleed) functions, and there are
gauze sieves under the primaries to exclude dust.
Each playing piston has a thin cloth washer underneath which acts as
a bottom stop buffer and was probably intended to be a bore wiper and
dust seal. These seem to have shrunk a little on my example, but
anyway none of the compressed graphite pistons was stuck or badly
scored. I've attached a drawing showing the mechanism enlarged and
actual size which I hope will make its operation clear.
There are many other fine construction points which I will detail
sometime later. The main reason the Pistonola was unsuccessful must
surely have been that it was in competition with the big boys and was
never fitted into a famous name piano. And it does seem to have
suffered a particular assembly problem: paper-faced hard rubber gaskets
which were fitted between the valve blocks and cylinder heads. Over
times these exuded as miniature doughnuts into the 3/32" drillings,
partially or completely blocking them with obvious results, especially
if any heads were over-tightened during manufacture or servicing.
Somewhere I have a review of the Pistonola by a well-known music critic
who noted how efficient and responsive it was. And the Boyd piano
models into which Pistonolas were fitted are reasonably well-scaled and
well-made instruments. So don't write them off too soon.
Wivenhoe, Essex, UK.
[ Thanks for the fine article, Paddy. I'll place the drawing image
[ and your text at http://mmd.foxtail.com/Pictures/Pistonola/
[ -- Robbie