[pianotech] snugging up an agraffe

[Ed Foote]

 Greetings, 
  I wrote: 
" The threads would strip before I could snug them sufficiently in position." 

Andrew asks: 

 >>seems to me that you would file the bottom of the agraffe or countersink a little to have an agraffe snug up at the desired angle to the strings.  Am I missing something here?  pushing brass to the point of stripping the threads seems to miss a step... <<

     How much is too much when seating the agraffe? I bend metal in the last few degrees.   I think I can feel the metal in the tapered bottom skirt give a little bit when I apply the final squish. Since it is tapered, there is an initial give that rapidly grows stiffer. With some practice, I can usually get them to seat right at the right angle at the precise moment they get "hard'. 
     Even when one alters the agraffe to fit, you still have to apply a certain amount of torque to seat it.   I sense torque with my hands when installing agraffes,and when the feel goes soft,(threads beginnng to bend away from the skirt), well before what I know to be a trustworthy and solid seating has occurred, I measure.  And yes, the undersize contributes to this problem, a lot.
     I like to put a small touch of grease or wax on the  threads, so that the top of the stem isn't being stressed by friction as I flatten the skirt to order. To the bestof my knowledge, none that I have installed have failed, so I feelpretty confident in my sense of feel for how they should behave.  I am not as fast as the factory guys, but the guy in the factory will never know if 10 years later, two or three times a day, agraffes finally let go in pianos he muscled into shape.  
When the agraffe comes to its final position, it needs to be as tight as possible without damaging the stem. If you have an old plate, break a few, get the feel for exceeding the limits, it is the best way to learn how much is too much.  
Regards, 
Ed Foote RPT 

 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ptg.org/pipermail/pianotech.php/attachments/20101004/cb883e98/attachment.htm>