Call for scaling spreadsheets

[Frank Emerson]

Although there are exceptions, it is a reasonable assumption that all
strings of a unison were designed to have the same tension.  I personally
disagree with the arguments of the few that intentionally vary the tension
within a unison.  For purposes of setting up a spreadsheet and graphing the
results, I treat all strings of a single bichord or trichord unison as
being identical, with respect to the speaking length.

Typically, a graph of tension will reveal a saw-tooth pattern, with a peak
at each point where the wire gage changes.  The goal is to determine wire
gages for each note that will yield the highest peak and the lowest dip at
equal distances from the target tension.  However, I have designed and
built two scales with all plain wire string have exactly the target
tension, 160 lb. throughout, and tapering slightly higher in the highest
octave.  I discontinued this practice, not because of any negative effects,
but because to was more trouble than it was worth, yielding no measurable
positive effects.

Frank Emerson
pianoguru at earthlink.net


> [Original Message]
> From: Jason Kanter <jkanter at rollingball.com>
> To: <pianoguru at earthlink.net>; Pianotech List <pianotech at ptg.org>
> Date: 9/30/2006 10:28:22 AM
> Subject: Re: Call for scaling spreadsheets
>
> Frank - do you shoot for equal tension per string, or equal tension per 
> unison?
>
> ----- Original Message ----- 
> From: "Frank Emerson" <pianoguru at earthlink.net>
> To: "Pianotech List" <pianotech at ptg.org>
> Sent: Saturday, September 30, 2006 12:05 AM
> Subject: Re: Call for scaling spreadsheets
>
>
> > Joe and Ron,
> >
> > I have been following this thread for some time.  As a designer of
scales
> > for new pianos, I could offer a VERY lengthy discourse on this subject.
I
> > do not have time to get into this, and on that scale.  Ron, your
> > contribution seems to me to be the most practical and applicable to the
> > immediate situation.  When one is designing a new piano, with a full 
> > palate
> > of variables at his disposal, many things must be taken into 
> > consideration.
> > On the other hand, when you have the bass strings, and every other 
> > variable
> > is "nailed down," all that you have to determine is the number of
strings
> > to replace with each wire gage.  Without control of any other variable,
> > such as speaking length, hitch pin placement, bridge configuration,
bridge
> > pin configuration, bridge design, soundboard design, rib placement and
> > design, etc., all you have to determine is the wire gage to use for each
> > note.
> >
> > Forget about inharmonicity, loudness factor, stiffness, characteristic
> > impedance, longitudinal mode partials, etc, etc.  All of these are valid
> > considerations if you are designing a new scale, but all are irrelevant,
> > and beyond your control, if all you need to know is what gages of wire
to
> > use for each note in the plain wire sections.  I have a multi-page
> > spreadsheet for such calculations, but it is an ever evolving thing. 
Each
> > time I use it, I add new calculations.  It would be useless to anyone
who
> > is not familiar with its evolutionary development.
> >
> > Count the hitch pins in each plain wire section.  This will tell you a
lot
> > about the original design intent. If there are three hitch pins for
every
> > two unisons, this simplifies things.  If there are more hitch pins than
> > this ratio, you have to determine how many strings have to be tied off
and
> > how the side bearing would be influenced.
> >
> > All you need, in terms of a spreadsheet, is a simple calculation of the
> > tension for each note and a graph to display the tension from note to 
> > note.
> > A reasonable target tension is 160psi.  Ideally you want to minimize the
> > range of fluctuation either side of this target.  Each time you increase
> > the diameter of the wire, you will see a peak in the graph.  This will
> > guide you in determining when and where to change the wire gage.  You
must
> > always have an even number of notes per wire gage, unless you have
enough
> > hitch pins to allow tied-off strings to account for this.  The goal is
to
> > keep the fluctuation in tension within as narrow a range as possible. 
In
> > smaller pianos, it is expected and acceptable to allow a slight upward
> > curve in the tension in the highest few notes of the treble.Get the 
> > tension
> > right in this range, and trust that the original design takes all of the
> > other considerations into account
> >
> > Frank Emerson
> > pianoguru at earthlink.net
> >
> >
> >> [Original Message]
> >> From: Ron Nossaman <rnossaman at cox.net>
> >> To: Pianotech List <pianotech at ptg.org>
> >> Date: 9/29/2006 3:37:59 PM
> >> Subject: Re: Call for scaling spreadsheets
> >>
> >>
> >> > I guess my question that started this great discussion was really can
> > one
> >> > start with a given ( the bass strings ) and produce a scale that will
> >> > compliment them.
> >> > Joe Goss RPT
> >>
> >> Sure one can, but it won't be all that complimentary because
> >> that's backward. Without moving bridges and changing speaking
> >> lengths, about all you can do with plain wire is determine
> >> tension ranges with wire gage. Inharmonicity will be what you
> >> get. By far the most control and improvement in rescaling
> >> happens in the bass, which in this case is already nailed down
> >> to whatever the existing bass strings give you.
> >> Ron N
> >>
> >
> > 
>
>