Greetings,  I have the equipment to measure valve transfer time,
in real time.  Of course the valve has to be in an operating piano
to obtain a valid measurement.  I have a Duo-Art, a Welte licensee
Vorsetzer and an Ampico B.  Otherwise, I would have to travel to the
piano with my stuff.

I can measure vacuum change at the entrance to the valve and the signal
tube in real time.  This is not easy to set up so I would only do it if
there is a compelling need that can not be met with ballpark figures
from the industry, casual measurement or theorized calculations.

The method involves measuring the vacuum at the signal tube with
a highly sensitive electronic vacuum gauge read by a computer data
logger.  The same is done at the inlet to the valve.  The two waveforms
are compared as holes pass over the tracker bar.  It is easy to see the
point at which the vacuum just starts to drop in the signal tube and
also the point at which it just starts to rise at the valve inlet.  The
difference is the transfer time.  Sort of like an oscilloscope for air.
In fact you might be able to do it with a scope if you have a steady
stream of holes that you can sync the scope to.

  [Later:] I got a letter from Robert Taylor to whom I sent the
  following response about valve transfer time.

Transfer time is a measure of time but the number you want (percentage
of hole required to effect transfer) can be derived from it.  Since the
paper is traveling at a fixed speed, the relationship between distance
(or exposed area of the hole) are directly related.

The following variables all effect the transfer time so you really need
a specific valve with carefully set adjustment to get a meaningful
figure and it would vary from one design to the next, of course:
bleed size, tracker bar port dimensions (only if narrower than the
hole), paper hole size, pouch well diameter and volume, pouch
clearance, valve seating diameter, valve button weight, pouch
stiffness, volume of signal tube and wall stiffness, operating vacuum
pressure and roll speed.

If you want the time to valve fully opening as opposed to just the
point at which it begins to open then you also need the valve travel.

If you really have a need for this and a specific valve that you want
tested, I can do it.  What is the need for such a figure?  Once you get
it, it will only apply to one specific valve design and setting.

My guess is that 50% sound like an average figure given all the delays
that influence it, unless the bleed is really small.

I have not thought out the details exactly but I think it might be
possible to derive the number you want by a specially made test roll
similar to the Welte test roll that had various hole sizes.  Just make
holes which have different area relationships to the original hole
(which hole size do you want, each manufacturer used two at least and
some three) in question.  Make one 20%, 30%, etc.  You can then pass
them over the tracker bar at various speeds and see what happens.

Best regards,
Spencer Chase
Laytonville, Calif.
http://www.spencerserolls.com/