In MMD 031207, Spencer Chase wrote:

> It was found that the theme regulator pressure decayed
> very slowly if the theme valves remained closed.
> Examination of coding on certain rolls suggests that the editors
> were well aware of this and exploited it for subtle effects.

I don't know about how the rolls editors used the slow decay of the
theme vacuum but I do have a story about this very problem.

About six years ago, I rebuilt a Duo-Art grand regulator which didn't
have the bleed hole for the theme vacuum.  I thought that it would be
best to seal the new theme pouch leather to make it as air tight as
possible to make them faster.  This was a mistake.

I also had problems finding the right leather thickness and stiffness
to use on the 2 flap valves used to keep the theme vacuum from going
into the accompaniment regulator.  I used 2 pieces for leather glued
together with PVC-E glue, which gave me the stiffness needed for this
application.  It also provided a greatly improved seal.

When testing the regulator in the piano, it was having problems with
the theme notes.  When the coding asked for a level of 14, it would
give this with no problem.  If the next theme note asked for intensity
5, it would play at level 14 on the very next note only, the next ones
would then play at the proper level 5.  In looking at how the regulator
was operating, the theme side was never collapsing back down in a
speedy fashion. Even though the accordion pneumatics were completely
relaxed.

My first thought was a leak in the theme knife valve.  I took the knife
blade out, tested it, and it sealed 100% when closed.  I went ahead and
resurfaced the knife valve a second and put everything back together.
The problem was still the same.  I repeated the same procedure on
accompaniment knife valve.  No change at all.  Must be a leak between
chambers or the top seal cloth not being glued properly.

I removed the top seal cloth on the regulator and could not find a way
to test for leaks between the 2 wood assemblies so I split the
regulator again.  Good thing I used leather between the wood surfaces,
otherwise this would have done damage to the parts.

I went ahead and checked all that I could for air tightness and
everything checked out great.  I cleaned every thing up and re-glued
the regulator back together and put it in the piano.  Guess what, the
problem changed slightly but was still there.

I found that if the vacuum motor were turned off, the theme regulator
just wouldn't relax until about 3-4 seconds later.   When it did start
to go, it would drop like a rock.

I am not proud about what my solution to the problem was but it is
reversible.  On the top seal cloth, there is a place that I put a small
1/32 to 1/16-inch hole that vented the theme regulator to the
atmosphere.  Everything worked right and the regulator problem was
fixed (but not the traditional way).  Let me explain.

What my real problem was that the theme regulator chambers didn't have
any leaks or paths for the vacuum to be released in the theme chamber.
The regulator releases this unwanted vacuum through the pouch leather
found on the theme valves.  It just so happens that the UK regulators
probably added a bleed to help eliminate this very problem.

You can test how much this leak is by connecting a hose to each of the
theme nipples and sucking. With these 2 theme lines being open to the
atmosphere when not activated, they provide the needed bleed to relieve
the unwanted vacuum for the next theme intensity change.

As Spencer Chase indicates, the roll editors may have exploited the use
of this extra vacuum in the Duo-Art theme regulator but I don't think
it would be the same from piano to piano.  It would depend on the
quality of the leather and would surly change as the leather ages.
Aeolian may have never sealed these pouches because of this very
problem.

Spencer wrote:

> Recently I performed some tests of vacuum levels while the
> Duo-Art was playing.  MIDI files were designed to test specific
> theories as to how the Duo-Art regulator worked.

I was wondering if this testing was done on a recently rebuilt regulator?
If the leather was recently replaced and/or sealed, it may not react
the same as if it were built 80 years ago when it was new.  We may
never know this answer unless there is documentation as to how the
leather was selected at the Aeolian factory for these theme pouches.

Pete Knobloch (Tempe Arizona, USA)