overcentering justified?

[A440A@aol.com]

Bob writes:

<< The tail arcing was done with the Spurlock jig.  Only the shank/flanges 
and hammers have been changed.  So, yes drop is changed. Balancier is same. They 
had a drop of maybe 3/8 below letoff on the old and I'm doing more like 1/16 
below let off for drop.  The old let off is about 1/32 in mid range.  << 

OK, there are more changes in this action than just hammers, so the 
difference in repetition speed may arise from several reasons.  A drop of 3/8" is way 
excessive, the finger will be running into spring resistance far to early for 
maximum sensitivity at pianissimo levels. The friction between the jack and the 
knuckle at these softer levels will also be very high, since the balancier is 
not there to help support the weight,(which it is doing with sliding friction 
of its own as the action enters the escapement zone).  
   I had posted earlier on my preference for as transparent an escapement as 
possible as a way of getting the maximum sensitivity out of an action.  This 
can be done with no loss of power or repetition speed.  It requires that the 
jack be set as far proximal as possible without losing power on the FFF blows, 
the weight of the hammer and shank being supported as much as possible by the 
balancier,(at rest), and using as soft a spring as possible with drop and 
let-off being one and the same distance from the string.  Often, when this setup is 
finished, the jack tender and drop pad will contact their respective stops 
simultaneously, but that particular result isn't necessary to produce a very 
controllable escapement.    
   There is a string of events that must be considered.  The keydip must be 
set before the drop is finalized. The keydip can't be determined,(in the 
aftertouch priority method I use), before the hammer blow and let-off is set. The 
hammer blow can't be set without knowing what sort of geometry you have extant 
in the action, because you don't want to be forced into an extremely shallow or 
deep keydip just to accommodate a particular blow distance.These things all 
have to work together.  
   On a Steinway, I begin with a .390" keydip, appropriate let-off, and see 
what blow is required to achieve a .030" aftertouch.  You don't want the hammer 
shanks to be farther than 1/4" from the rest cushions, nor resting on them.  
If I have to fudge things around, such as when the blow comes out to be 1 7/8" 
or more, I can go .400" on the keydip with .020" aftertouch in there.  On the 
occasional D that turns up wanting a 1 3/4" blow, maybe a .375" dip with 
.040" aftertouch.  As Roger Jolly has pointed out, aftertouch is anathema to 
repetition speed, since the key must travel up through it before anything begins to 
reset.    


>>If the jack is not resetting under the knuckle during fast repetition, is 
the knuckle being sent flying up again from contact with the top of the rep. 
lever ?  I decided bouncing was not a good word since it implies that one or 
both of the parts are being set into motion through a compression and release of 
energy from their surface.  I'm sure the knuckle surface compacts or 
compresses but not in the same sense as a basketball.   >>

    Try this experiment: with a straight arm, and using the middle joint of 
one of your fingers, play a note by coming straight down on the key.  You will 
hear the hammer bobble against the string!  It sounds exactly like too close 
of a let-off, but what is really happening is that the compression being 
released in the action train is actually throwing the hammer off the top of the jack 
before escapement. Your knuckle is too hard to accept the reverse wave of 
energy that the action responds with.  However, playing normally, (via the pad of 
the finger on the key), the key's recoil is absorbed sufficiently to reduce 
the shock wave and keep the action entrained throughout its movement.  So, yes, 
the compaction in the action train, be it knuckle or capstan felt, or 
bushings or spring in the key, etc., does absorb power. 
   What does this have to do with repetition speed?  Well, it depends on what 
sort of speed we are discussing.  Consider the hammer at check, how fast can 
you play another note?  This will depend entirely on how long it takes the key 
to rise and allow the jack to reset. Excessively high checking will increase 
the speed, but decrease the power,(how much power can you get from a 1/4" 
hammer blow?) .    This may not be the same thing as playing as many notes as fast 
as possible,(think Ravel), in which the action is never allowed to come to a 
complete rest.  In the first case, checking height is everything.  In the 
second, there are numerous causes for loss of repetition.  One of them is the jack 
pinning, which if too loose will allow the jack to oscillate out of control 
upon escapement, thereby not being in position to reset, even if everything 
else if correct.    
   I am not clear on why the new parts are repeating more slowly than the 
old, but I suspect that there is a difference in knuckle diameter, or perhaps 
there is less room for the jack in the balancier slot at check.  If the jack is 
trapped between the knuckle and the mortise stop pad at check, it will cause a 
lot of friction between the jack and the proximal side of the knuckle. This 
greatly decreases the initial movement of the action's return to resetting the 
jack, and could be where you are losing some repetition speed.  
   Let us know exactly what is the condition of the action at check and there 
may be some other things that suggest themselves.  I am assuming that all the 
pinning and friction ,(particularly the knuckle/balancier) is correct.  
Regards,  
Ed Foote RPT 
www.uk-piano.org/edfoote/
www.uk-piano.org/edfoote/well_tempered_piano.html
 <A HREF="http://artists.mp3s.com/artists/399/six_degrees_of_tonality.html">
MP3.com: Six Degrees of Tonality</A>