[CAUT] Sperrhake Harpsichord wire

[Ed Sutton]

Hubbard, on p. 281, in a footnote, states that theoretically strings of 
equal length and the same material should all break at the same pitch, 
regardless of diameter, then goes on: "The process of drawing steel into 
fine wires seems to produce a structure composed of a hard sheath around a 
softer core. This sheath is always roughly equal in thickness, and thus 
occupies a larger percentage of the diameter of a thin string than of a 
thick one. Therefore a thin string will actually stand a slightly higher 
pitch than a thick one."

He is, of course, refering to practical experience as a harpsichord builder 
stringing the high end of the instrument.

The explanation seems reasonable to me.

Ed Sutton

----- Original Message ----- 
From: "Fred Sturm" <fssturm at unm.edu>
To: <caut at ptg.org>
Sent: Saturday, December 05, 2009 7:54 PM
Subject: Re: [CAUT] Sperrhake Harpsichord wire


> On Dec 5, 2009, at 1:56 PM, Ron Nossaman wrote:
>
>> In that REAL WORLD you were so insistent on, are you likely to  deviate 
>> more than 0.001" on that string replacement? And wouldn't  that 
>> calculated break% difference be at worst, a half percent by  your 
>> formulas? And how, by any reasonable criteria, is this worth  the time 
>> spent on this thread trying to drag useful information out  of it?
>
>
> When you are in the real world with a string breakage problem, you  are 
> probably in a high range of the breaking %, no? What are you going  to do? 
> Is changing the speaking length an option? Not often. So you  need to have 
> a knowledge of the principle that underlies this. Try  entering lengths in 
> your spreadsheet that get you in the 55-60% break  range. Experiment with 
> increasing and decreasing diameter. You'll find  (or at least I have 
> found) that a .001" change narrower can get you  1-2% lower break%. A 
> change wider gets you higher break%. It's a  consistent pattern. I grant 
> you, it isn't dramatic, and might not turn  out to be significant on 
> pianos, in real application. But it is clear  that if you are going to 
> experiment with substituting string size, to  deal with a note that is 
> breaking consistently, a smaller size is much  more likely to solve your 
> problem. I haven't found an instance where  the pattern was reversed, have 
> you?
> I have never done this on pianos (except that if I happen to be  missing 
> the half size at the moment, I go small rather than large). My  experience 
> in this is with harpsichords. On harpsichords, I suspect  that the break% 
> variance is probably larger in proportion, judging  from my experience. 
> Partly because we are dealing with proportionally  larger differences of 
> string mass (.014 versus .015 is a bigger  difference than .040 to .041), 
> and often with scales that are a lot  more "ambitious" (pushing the break% 
> envelope). Maybe someone on the  list has a set up for harpsichord 
> tension, break%, etc and can run  some figures or provide a spreadsheet.
> I do know from several experiences of trying, that, for example, 
> replacing the .015 brass string that is breaking with .016 is a waste  of 
> time and material. It will break, probably faster than the .015  (probably 
> when you get to about 1/4 step flat of target pitch).  Replacing it with 
> .014 is pretty likely to be successful.
> Which is really the only point I was offering, except that I wanted  to 
> explain why it is so, because it is counter-intuitive. You are  right, far 
> more smoke than fire in this thread, especially in  application to pianos. 
> But it is a useful thing to know if you work on  harpsichords, where this 
> problem often arises. And that is what this  thread was about originally.
> I don't believe, though, that any of the statements I made about  pianos 
> was wrong (though the significance is maybe lower than the  verbiage would 
> indicate). For instance, I believe you will find it is  true that any 
> string on a piano replaced with 13 gauge will have an  equal or lower 
> break% than with its existing gauge (and similarly for  other gauges 
> moving down the scale).
> Regards,
> Fred Sturm
> University of New Mexico
> fssturm at unm.edu
>
>
>
>
>