Sam Harris' problem with low power on rewind reminded me of a solution
I came up with about low power rewinds.  With the # 2800 automated
suction box kit from PPCo, you rely on the same box as the # 650, but
receive an added rewind pneumatic and two valves for a fast dump.  That
system has two basic problems.

1.  The suction box should have two independent controls -- one for
rewind, and one for play, and,

2.  In most cases, the pneumatic is too large with too much volume to
exhaust for two little pneumatic valves, so it isn't nearly quick
enough.

With a tight player, the motor is just "growling" along.  At that low
power, there is no way it can ever throw a large bellows fully into
rewind.  Also, with a tight player, once you have set the suction motor
speed for a comfortable listening level, it often will not start up
again at that level, anyway.  You have to give it a "boost," first,
then back it down again, each time.

I know two ways to do this pneumatically.

1.  When (and only when) the suction box plays the player, an
adjustable dump valve opens (via an electric pipe valve) and stays open,
dumping a large portion of the play vacuum into a felt muffler through
a slot.  As soon as the rewind slot in the paper passes the bar, That
electric valve is shut off and all the force of the play vacuum can go
to the rewind pneumatic, which will then work every time.  By making an
adjustable dump valve using a slide cover, you can fine-tune the leak.

2.  If, by testing the rewind bellows, you are able to make it reverse
at that play pressure, if only you had a little better flow
characteristics than two Wurlitzer valve blocks can give you, then you
should make yourself a "dump pouch valve." In the Orchestrion Builder's
Manual and Pneumatics Handbook," I give full plans for it.  See the
"Crash Cymbal" plans.

As far as the independent controls on the suction box, you can do that
yourself, if you know something about triac motor control circuits, but
you'll need 3 wires going out to the microswitch instead of 2.  The
microswitch then will be able to short across potentiometer #1 or
conversely #2 in the gate bias circuit, as I recall, instead of using
both of them during play, and one of them during reroll, as it is right
now.

To show you how ridiculous this circuit really is in its present form,
I have had pumper players so easy to pump that I had to turn the play
control all the way off, and then turn down the reroll control so low
that it still didn't have enough power on reroll to move the rewind.
The mechanical advantages of player transmissions are different, so
some will work -- sort of, and others won't work at all.

A 3-wire control to the microswitch is all PPCo would need to have
a good box.  I suggested this to the fellow who builds them, but
unfortunately, he doesn't think I know too much.  He told me he's been
building them this way for (umpteen) years.  He hasn't had any
complaints yet, and he's not about to change now.

Besides, he says, I don't even appreciate the "problems" he has had to
solve over the years as supplies become unavailable.  (Translation:
the cheap surplus parts get used up.)  His circuit is still basically
the same one featured in the 1950 RCA Transistor Manual.  No wonder he
has "problems."  But he will not listen to solutions, either.  He made
it clear that he didn't appreciate my concerns.

Craig Brougher

 [ Often those guys with no complaints end up with no business!
 [ The successful supplier doesn't wait for complaints -- he calls
 [ upon his customers and _listens_.  -- Robbie