[CAUT] moving capstans question

[Fred Sturm]

Hi Paul,
	What you are describing is a convergence issue, rather than a  
leverage issue (though there may be leverage issues as well). The  
line from balance hole bottom to wipp center is the "line of  
convergence" and you want the top of the capstan/bottom of the wipp  
to be on the line when the key is halfway through a keystroke. This  
mostly affects friction, how much the capstan and wipp cushion felt  
rub back and forth against each other. The least amount will occur  
when you have it set up as I described. It usually doesn't have that  
much of a noticeable affect on touch unless the capstans and wipp  
cushions are "gunked up." It will affect wear, how long the cushion  
felt lasts and how soon it becomes badly cupped.
	If this is enough of a problem to address, the best way to address  
it is with wipp heel changes. For what you describe, you want a  
thinner wipp heel. Picture it like this: hold everything in place,  
and shave X amount off the bottom of the wipp heel. You will now have  
to raise the capstan to meet it. Hence, the contact between capstan  
and wipp cushion will be higher relative to the line of convergence.
	Moving the capstan line will affect convergence, but only "by the  
way." Let's be clear what we mean by back and forward: back, toward  
the back of the key, forward toward the front of the key and the  
player. Moving back will make the capstan lower relative to the line  
of convergence. Moving it forward will make it higher. (The line of  
convergence is slanted relative to the top of the key). But the main  
effect is on leverage. Moving the capstan back means higher ratio,  
and heavier touch (all other things being equal), and means less  
keydip for the same blow. Moving forward does the opposite: lower  
ratio, lighter touch, and more keydip for the same blow.
	Raising or lowering the stack has a lot of side effects, especially  
having to do with hammer bore distance and how squarely the hammer  
hits the string. You don't want to go there unless you are able to  
think of all the effects. Well, raising it is simple to experiment  
with, just a few shims you can easily remove, and not a lot of time  
lost. And you can consider just shimming the back legs, which you can  
do more of without causing problems getting the action assembly in  
and out under the pinblock. But it does have other regulation effects  
besides just the convergence thing.
	I guess I would start by questioning whether convergence is the  
issue here that is making people hate the piano. The most likely  
problem is new heavy hammers on old knuckle geometry (16 mm or so vs.  
17 mm or so current specs). In that case, best solution is probably  
moving the knuckles, whether that means new shanks with different  
knuckle placement, or cutting new slots on the old and installing new  
knuckles - which requires specialty jigs.
Regards,
Fred Sturm
University of New Mexico
fssturm at unm.edu



On Sep 21, 2007, at 8:34 AM, Paul T Williams wrote:

>
> Hi List!
>
> Happy Friday!
>
> I am working on a 70's Sty B with a horrible action geometry  
> problem. All the students and the professor hate playing it and I  
> don't blame them! The line from key balance hole to wip has the  
> capstan way too low making the DW close to 70g throughout.  It has  
> a renner action and hammers.  the wips and capstans are such that I  
> could relocate the capstans back by about 3-4mm.  Would that make a  
> big enough difference in the DW? It won't affect the line.  I  
> looked at raising the stack, but there isn't enough room to raise  
> it more than 1-2mm.  Should I try both?
>
> Also, is the measurement in Steinways made from center of balance  
> rail hole on the top of the key, or from the back of the key?  This  
> is a new endeavor for me and I don't want to make things worse.
>
> Still green in Lincoln,
>
> Paul

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/caut.php/attachments/20070921/6c5a34ce/attachment.html