[pianotech] capstain/wippen angle, was: key position at rest

[Mike Spalding]

Nick Gravagne wrote:
>
> * I believe that the explanation was that the slide path, and not 
> necessarily the capstan top or the whip heal, follows the involute 
> curve. This slide path would have to be precisely drawn out, plotted, 
> graphed, extrapolated and formularized in order to evaluate what kind 
> of a curve it really describes.*
>
In gears, it's the contact surface of the gear which is machined to an 
involute curve. The path followed by the contact point (actually a 
series of contact points as the teeth slide over each other) is a 
straight line. If it doesn't follow a straight line, it's not involute 
gearing. I don't know, offhand, how you would arrange to have the path 
of the contact point follow an involute path, nor do I think there would 
be any special magic in the results.
>
> * *
>
> * *
>
> *I think that the power transmission idea holds more promise, but I 
> would like to investigate the involute as it too (or whatever curve 
> exists) is a clue to the power acceleration curve, if there is one.*
>
Here's why I am attracted to Dale's explanation. Picture the key and 
wippen, from the side. At rest, the angled capstain contacts the wippen 
at the front edge. Calculate the action ratio. Now slowly depress the 
key and watch the contact point move towards the rear edge of the 
capstain due to the rotation of the key and wippen. How has the action 
ratio changed? What does this feel like to the pianist? How will it 
behave during weigh-off?

*
*

> *This all may be relatively useless for the working tech, but since 
> the question routinely arises RE “what should do with these angled 
> capstans and sloping whip heal?” perhaps we can come up with a better 
> answer.*
>
I think it will be very useful to this working tech, to have had my eyes 
opened to the positive aspects of the sloping capstain and heal design. 
Unfortunately, this was not the first time that I rejected and changed a 
traditional design simply because I didn't fully understand it.
>
> **
>
> * *
>
> *Perhaps this has been done already. Any data out there?*
>
> * *
>
> *Thanks again, Mike and Dale*
>
> * *
>
> *NG*
>