[CAUT] stack fit to keyframe

[John Delacour]

At 11:19 pm -0400 11/9/06, A440A at aol.com wrote:

>Bob writes:
>
>>  >>One can't argue that the principles
>>calling for the magic line are correct, but in reality I've never been able
>>to detect an increase in friction in an action that deviates from the magic
>>line.<<
>>
>
>Nor I.  I am not sure what the "principles'  are, though.  Allowing 
>the capstan to pass above the magic line before escapement is 
>something I try to avoid, since it begins sending the ration back 
>"down" just as you can use the increased leverage to keep up the 
>hammer velocity.(I posted on this last month).
>   Like Bob, I have varied the settings to see what happens to 
>friction when the magic line is ignored, and I couldn't find any 
>difference between the possible settings in a given action.

I'll avoid the use of "magic line" and refer instead to touching 
arcs.  As regards the contact between the pilot and the lever heel, 
whether the arc from the lever centre touches the arc from the 
key-balance before or after set-off has a negligeable effect on 
hammer velocity owing to the relatively large radiuses involved and 
the relatively small deviation from a straight line.  Far more 
significant in this regard are the arcs from lever-centre to fly and 
from hammer-centre to fly.

As to potential friction at the pilot/lever-heel contact area, this 
will be most significant at the start of the stroke, when movement of 
the system is initiated.  After that, additional pressure will be 
applied to this contact area during the set-off procedures, but 
whatever the degree of this pressure it will be less significant 
since the profiles are already in motion in relation to each other. 
To have the arcs touch when the system is at rest in order to 
minimise the friction at this moment requires the action to be 
slightly higher and also increases pilot friction at the moment of 
set-off.  In practice a touching of the arcs at some point between 
half-blow and set-off is what a well-designed action seems to require 
and this happens to be very close to the much-trumpeted "magic line", 
but experience also shows that owing to the smoothness of the 
polished brass to cloth/buckskin contact and the very small amount of 
relative movement between the profiles, wear and noise due to 
friction are almost insignificant for minor deviations from the norm. 
That said, any good action is designed according to very precise 
measurements and geometry.  To change the point at which arcs touch 
in any part of the action will affect other touching arcs and a 
supposed idealization of one point in the action can easily introduce 
significant errors elsewhere to negate the improvement or worse.

JD