<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> <title></title> </head> <body bgcolor="#ffffff" text="#000000"> Hi Ed Comments interspersed below: <blockquote>    Any action will fail to repeat if the second note is struck before the key resets under the knuckle.  There is a minimun distance the key must be allowed to rise before the reset.  If the key is struck before this minimum distance is achieved, the jack will either skip or will not touch the knuckle, at all.  On a normal action, this is just a one time failure and as soon as the key is allowed up, the note will once again repeat. </blockquote> Yes... and on any well regulated grand action I know of... this distance is well less then the full return of the key to its rest position. <blockquote>       On an action in which the hammershank cushion is too far below the hammer's resting position, the skip, (cited above) allows the hammershank to fall so far that the capstan is holding the whippen too high for the jack to reset.  In other words, the knuckle is below the lowest jack height possible.   Hence, a note that doesn't play until the hammershank is bounce up into the jack's range.          ( I hope that makes sense...)  </blockquote> Sure it makes sense.... except that the skip shouldn't happen if the hammer is allowed to fall that far.  Even in the condition where the hammer isnt checked because of very soft play.  Even if the jack does skip and the capstan holds the jack top higher then the knuckle position... you should be able to just regulate the capstan low enough to alleviate that condition. This should worst case result in non standard key dip.   <blockquote> <blockquote><<  The rep lever is supposed to lift the hammer via the knuckle up to a position where the jack can slip under as soon as (nearly immediatly after) the key is released.  >> </blockquote> </blockquote> <blockquote>       I must respectfully disagree.  The purpose of the rep lever is to increase the speed of the key return.  We measure the spring strength by watching the hammer rise, but in rapid repetition, that event doesn't occur.  Under fast repetition,  the hammer shank doesn't rise between blows, but rather, stays at checking distance while the key and whippen, (under the influence of the spring), return to their "reset" position.  One can ascertain this for themselves by playing middle C so that the hammer is held in check, and then slowly depressing B3 so that the hammer is at drop.  Simultaneously, release both keys and observe that the C return far, far more rapidly than the B. </blockquote> While I'll agree that the rise doesn't occur during usual play (at least not with full key release and moderate spring and key mass levels), I do not agree that this is synonymous with stating the purpose of the rep lever is to increase the speed of key return.   The purpose of the rep lever is to allow the action to return to firing position before the released key is fully returned.  If the jack isnt under the knuckle well before that...then something is clearly wrong.  Every action I regulate allows for fast rapid repetition without the key anywhere near full return. This can easily be demonstrated by artifically limiting the height of key return to say 3/4 full return. You should still be able to play quick repeats. <blockquote>      One can also see, on an action model, that when the key is released while the hammer is held in check, the hammer doesn't rise, but rather, drops to its at rest position. Under high speed conditions, there is no time for the hammer to rise between notes. </blockquote> This happens because when we test rep spring we hold the key down and barely release pressure so as to see how much rise we can induce... its a test.  But in real play the same forces are are work and these forces do the same thing relative to the parts as under the test... the only difference is that some of that force is allowed to push downwards on the key and the other half of the whippen... the end result should be pretty much the same. <blockquote>     This is why there is no difference in repetition speed between a spring that slams the hammer off the jack when released and a spring that simple lifts the hammer as fast as possible without being felt in the key. </blockquote> I must disagree here as well.  First and formost because it is nearly impossible to not feel the rep lever pushing against the key if you have any real pushing strength in the spring.  And because the difference between that extreme and even reasonable firm lift using the standard test, barely releasing the key out of check, does indeed result in a difference in repetition speed. <blockquote>  Unfortunately, many regulators destroy the sensitivity of escapement by pursuing the former course in the mistaken belief that it improves repetition.   Checking height has far more effect on repetition speed than anything else. </blockquote> I'll agree that there is a limit to how fast the action can be made to play... and how strong the rep lever should be felt through the key... but this is a bit more complicated then just so... involving relative tightness of jack and rep lever centers... hammer flange centers, and for that matter whippen flange centers and key bushing friction.  Checking height has likewise limits....   All this said... it is easy enough to regulate an action so that it plays faster then then ability of any human to play. <blockquote>Regards, Ed Foote RPT </blockquote> I cant say I've ever run into this condition perse'.  I've seen things somewhat similar... but these are always regulation issues.  I still don't see how relative positions of the center pins comes into all this.... tho I am sure someone out there can explain this to me. Cheers RicB </body> </html>