Some may take exception to my letter suggesting that the Ampico B
tracking fingers do not at all contribute to the tearing of rolls.
That's fine with me. I think meaningful discussions and differences of
opinion are healthy, and I'll try to answer the disagreement here.

Since I've been told that tracking problems in the B can't be caused by
stiff tracking ears, let me explain my reasoning further and see if it
doesn't make complete sense. Then the reader can make up his own mind
who's right, and who's wrong.

The tracking ears are sprung with flat steel springs which are just
too strong. When you bend one, you bend both, because they are
connected between fingers with a captured floating valve wire
(tracker rod) that works as soon as it gets in proximity to one or
the other sensing nipple. All Ampico Bs are too stiff to actuate on
weak, feathered roll edges (sorry, but that's just a fact), and even
though the roll may still play, tracking off to the side of the
feathered edge, it tears up the roll on reroll because the paper
favors one side of the take-up spool while the tracking pneumatic
centers itself by the bellows return spring.

It is normal in the case of all reproducing pianos that the reroll
brake is stronger than in the case of an 88 note player because they
were designed to be automatic and to handle the large rolls as one
finds in classical music. The Ampico B reroll brake is inoperative on
rolls shorter than 40 feet or so, and never operates at all. However,
it still has a certain minimum residual drag on all rolls.

As far as the reroll snubbing is concerned, it is momentary to stop
the paper and tighten up large rolls. The spool clutch is then
adjusted such that there is no undue tension on the drive motor or the
roll. I have never seen the roll tightening feature actually tear the
roll, although I can see how that would rarely happen, but remember,
the roll is stationary when the clutch slips, timed by a cam-operated
pallet on the electric motor.

Normally, the roll can usually only tear when it's in motion, and only
then if it is being rewound crooked -- over against one of the roll
flanges. That will happen if the tracking fingers are sprung too hard
against weak feathered edge paper.

Ampico says this in their manual: "By bringing the take-up spool to a
dead stop periodically the edges of the paper are not damaged by the
flanges of the music roll spool." That is correct.

As far as never wanting to change tracker finger tension, and as far
as one technician's promise that it will never work-- well, I have
modified I think 13 or 14 of the restored Ampico model B's this way,
and they have all worked perfectly. My own has been tracking the
absolute worst rolls you ever saw, and it will not tear them because
the fingers are light enough to operate on feathered edges. These
model Bs no longer tear rolls, and the only thing I changed was
finger tension, never brake tension.

Now does it make sense?

Regarding the "oscillation" that  is experienced when the fingers are
lightened-- that's easy to fix. One may have tried this at one time and
lightened them too much, and/or shortened the gap between the sensing
nipples too much. Each individual Ampico B tracker requires its own
optimum gap adjust. The book says to set with a 1/32" feeler gauge,
but that's never worked quite as good as I can get them. It's also a
good idea to start by cleaning off the surfaces that approach, since
corrosion and dirt affects the way it works. The fuzzier the surfaces,
the more "oscillation" you would get.

Some people also think they have to put leather pads on the valve wire
ends.  That isn't necessary, and will cause the oscillation, all by
itself.  Oscillation is caused by feedback, and feedback is caused by
too little hysteresis dead band, created by the proper gap. Once you
have tested the pneumatics and know they are good and tight with no
leaks, only then will your other adjustments work the way they are
supposed to.

Craig Brougher