Nicholas Simons wrote in 040429 MMDigest:


>> Problem: In a hand-cranked street organ, obscenely called 'monkey
>> organs' in USA, one is providing a constant flow rate of air all the
>> time.  The feeders are designed to supply the maximum potentially
>> demanded airflow at all times, although some very short high demands
>> may be able to be catered for by the reservoir, but in a portable
>> machine the size of the reservoir is limited.  The paradox is that
>> when the music is at its quietest the spill is at its loudest, but
>> more importantly for the operator, he is wasting all that muscle power.


I agree that what an electronic engineer calls "shunt regulation" is
inefficient and a waste of energy.  Also I think "hurdy-gurdy" is more
"obscene", being wrong.  In the USA one cannot take an organ out in
public without being asked, "Where's your monkey?!"


>> The challenge is to design a simple mechanical linkage, within the
>> confines of the existing organ case, that varies the stroke of the
>> feeders in accordance with the demands of the music.  The result
>> being that the crank torque will also vary, and there will be no
>> wasted effort.


First, unless we are talking about a big Dutch-sized street organ,
and perhaps even in that case, in is not a good idea to have the crank
torque vary.  Experience with my 20er and 31er crank organs shows that
a steady reaction torque is needed to help the operator keep the crank
RPM constant and steady.

Under the proposed system, soft passages would speed up, and suddenly
slog down when a big chord progression came along.  A geared-up
flywheel could help with this problem, however.

Now to the design.  Probably the simplest feedback mechanism would be
something that defeats (props open) the intake valves in the feeders
for part of the stroke.  That is, when a feeder first starts to close
and compress its load of air, the flap valve that normally slaps shut
to prevent the air from escaping back into the atmosphere, instead is
propped open to return some of that volume of air back to the
atmosphere.

Other than some air friction in sucking that air in and back out, there
would be no work done on that portion of the air, since it doesn't get
compressed.

When the reservoir top gets to where it would trip the spill valve,
it could instead prop the feeder flap valves open.  Getting the linkage
to these valves may be tricky, since in the standard French feeder they
are in the moving center board, but then engineering should be fun,
right?

Johan Liljencrants, a Swedish builder of a large hobbyist band organ,
uses pneumatic relays to operate his feeder valves, so his system may
be easy to adapt to such feedback.  Knowing Johan, he probably has
already!

This would, by the way, make for a sudden, jerky loss of torque on the
crank handle, and personally I don't know that I'd want it.  But it's
a start.  Johan's organ is motor-driven, so it saves electricity, not
your arm.


>> If correctly designed, there will be no need for a spill valve.


I agree, though a spill should be retained in case the feedback linkage
fails.

One might consider gearing the crank up to a centrifugal blower and
using a series (gate valve) regulator such as on church pipe organs.

A related problem, the opposite or "dual", is how to most efficiently
drive an organ from a spring or falling weight.  Turns out that a
series regulator will almost stop the feeder motion cold when no wind
is needed, but the feeder crank will speed up as needed when the
reservoir is not full.


>> Have fun, ...


Oh yes, I am.  Thanks, Mike Knudsen

 [ Hans van Oost contributed to the MMD Archives a short article and
 [ drawing showing an internal spill valve which conserves energy.
 [ The design is adapted from Gavioli.  See "Bellows Pump with
 [ Internal Shunt Regulator" at http://mmd.foxtail.com/Tech/
 [ -- Robbie