Amica Ampico Article: Fourth installment commenting on technical errors.

So far, we have quoted Jeffrey Morgan to show that :

1. He was unaware that the Ampico B system compensated for stack
   transients.

2. He didn't really understand the purpose of the stack equalizer
   pneumatics.

3. He felt that a particularly overcomplicated, "cumbersome and slow" (his
   words) design was extremely ingenious.

4. He thought that the first intensity set point of the Crescendo
   pneumatic was somehow fed back to it from the stack.

5. He totally mischaracterizes the purpose and use of crescendos and
   believes that since the Model A crescendos are independent of the
   intensities while the Model B amplifies the intensities and is
   therefore "dependent on the intensities" that proves the Model B is
   incompatible with Model A rolls.

6. He believes that binary coding on the roll controls the crescendos

   I might add one clarification to his first point--"the Model B
crescendo system does not initiate corrective action for stack
transients as do most earlier Ampico crescendo systems."

   I take that to mean the expression system as a whole, and not the
crescendo system only.  Because in neither piano does the crescendo have
to take corrective action for stack transients at all.  That is why each
piano's crescendo is powered directly by the pump.

   And if he is referring to the action of the amplifier through the
Model A crescendo, then he has a double problem, because he is either
admitting that when the Modify Switch is set to Normal or Medium, (both
are the same) and the amplifier switched out of the circuit, by his own
criteria, it becomes a direct-to the-pump-spill crescendo, just like
the Model B.  So either way, it is wrong.  And if he believes that the
crescendo is required to also correct for stack transients, he would be
inventing another system altogether which would not even play on an
Ampico.  So even supposing that he meant the expression regulation
system when  he said "crescendo system" is giving him the benefit of
the doubt.  Otherwise, he would be wrong about both the Model A and B.

    Now to the next points: (can you believe this?)

7. "The Model B crescendo does not operate in the same four modes as
   previously described...  though at faster speeds.  It also employs the
   same binary language using two trackerbar ports." p 71, para 4.

ANS: We have already covered the fallacy of thinking that this is a
binary code-controlled device.  We have explained that the Ampico
Crescendo is actually an analog ramping device, not a binary
controller.  See Inspector's Instruction Book, 1919, pp 4-6.

   The fact that the B Crescendo moves at twice the speed of the Model
A is calculated, since the precondition given Dr. Hickman was that the
new piano had to be fully compatible with the 1921-22 standard
dictating roll coding format.  Any extra capability could not compromise
its ability to play at least as well, and usually better, the older
Model A rolls.  It succeeds wonderfully.  Any Model B that does not meet
this test is either not set up right, or has deficiencies in its
restoration -- usually valves and ball bleeds.

   The reason that the Model B crescendo collapses twice as fast as the
Model A is because its effect by percentage of its closure is only half
as much! Now does it make sense? And you know, this isn't an opinion.
It is provable.  I have measured it and know what I'm talking about.
Granted, there is an exponential function involved, so it isn't a
perfectly linear relationship, but more than adequate.  It fully
approximates the Model A, if that was ever the criteria anyway.

   Knowing Dr. Hickman, he was thorough and had a very mathematical
mind.  He would have simply calculated it from the ideal roll coding
technique, whereas Charles Stoddard  was a seat-of-the-pants inventor.
He would not have calculated anything, probably. He would have built a
testing apparatus to test the response and get what he wanted that
way.

   The purpose of telling us that the Model B does not do the same
things a Model A does is to cast aspersions on its design and to call
it incompatible.  This is not true.   The criteria was not to analogize
the Model A, but to build a better player which would be designed
around the roll coding format.  Everybody at Ampico agreed, including
the eternal skeptic, Angelico Valerio who was frankly amazed.

8. "In the second amplification position, the crescendo capability is
   nullified." p 71, para 6.

ANS: Here is the most ridiculous, embarrassing statement in the entire
article.  Jeffrey just wrote without thinking, I would hope.  This is
exactly the same thing as saying, "When the Model A crescendo is fully
closed, its crescendo capability is nullified." Neither statement makes
any sense whatever, because in the fully ON position, both are fully
ON.  There is no more crescendo to be had, so the crescendo is
"nullified." Wwell, that's apparently acceptable for the Ampico "A",
but inexcusable for the Model B.

   And where, perchance, are the technical committee advisors on this
one?  Apparently they didn't know any better, themselves.  "That's real
technical, complicated stuff there.  Only a engunear could figger it
out.  That's jest over mah haid."

Craig B.