[pianotech] problem in mating hammers to strings

[George F Emerson]

While this discussion focuses on mating hammers to strings, it may be useful 
to consider how the hammers were mounted, with or without left to right tilt 
for clearance, and how the action was designed, in anticipation of hammer 
tilt, or not.  If there is ample spacing, note-to-note, tilting the hammers 
may not be necessary, although it might offer a little extra insurance 
against the eventual possibility of clearance problems.  On the other hand, 
if the note-to-note spacing is particularly tight, tilting the hammer may be 
unavoidable.  If the designer intends the hammers to be tilted, the “strike 
point scale” may be different from the action scale.  This is most evident 
at the bass-tenor break.  Since the bass hammers tilt to the right and the 
tenor to the left, the distance between the strike point of the last bass 
and first tenor will be less than the distance between their respective 
flange screws, a difference of about 2-4mm.

I am at a loss to explain in text how this works.  Next time you hang new 
hammers, dry fix the first three tenor hammers, with the centerline of the 
hammer perfectly vertical while the shank is horizontal.  Lift the middle 
hammer to its strike point, and it will move very close to the hammer to the 
left, even though they were perfectly spaced at rest.  Rotate all three 
hammers on their shanks to the left, just a tiny bit; it doesn’t take much. 
You can find a point at which they are still perfectly spaced at rest, while 
the middle hammer is evenly spaced between its neighbors, when elevated.

Some would argue that hammers must be hung perfectly vertical.  Other would 
argue that the bass and tenor hammers must be tilted.  I would suggest that 
we should first know what the design intent was.  There is not a lot that 
can be done about the note-to-note spacing.  We’re pretty much stuck with 
that.  If the action spacing is different from the strike points, as 
described above and suggesting an intent to tilt the hammers, but the hammer 
are hung with no tilt, the hammer flanges will need to be rotate to bring 
the strike point of the hammer under its strings (not an easy thing to do on 
tubular action rails).  Conversely, to tilt the hammers on an action that 
was not designed for it, presents the same problem in the opposite 
direction.  Then there is the misalignment of the knuckle to the wippen and 
the hammer tail to the backchecks.  This is not to say that you cannot 
violate the original design intent, if you feel you have good reason.  Just 
be aware of the unintended consequences.

All of that said, I suspect that David’s conjecture is correct.  On a very 
heavy blow, the hammer may well want to sway in the direction it is already 
tilting.  High speed photography does reveal an amazing amount of shank 
flexing, although such photograph usually draws attention to flexing in its 
vertical motion with little or no perspective to observe any sideways 
movement.  Pressing the hammer to the strings from the jack tender requires 
sufficient force to prevent the jack from tripping, and it is hard to judge 
just how much of that force is transferred from the hammer to the strings. 
I prefer to use a hook, made from music wire, to lift the shank and hammer 
to the strings.  At least one can better feel how much pressure is being 
applied from the hammer to the string, and control the pressure to be just 
enough to damp the string(s), not enough to encourage the hammer to flex in 
the direction it is already tilting.  This could be used to support the 
argument for not tilting hammers (although I generally do).  But, It may 
just have to be a trade off that has to be made in a particularly tight 
action scale.


Frank Emerson

----- Original Message ----- 
From: "David Nereson" <da88ve at gmail.com>
To: <pianotech at ptg.org>
Sent: Tuesday, March 01, 2011 9:33 PM
Subject: [pianotech] problem in mating hammers to strings


> I've encountered this several times now with grand pianos.  I level 
> strings.  Then for each unison I hold the hammer up against the strings by 
> pressing upward on the jack tender, and pluck the strings.  The ones that 
> ring aren't being damped by the hammer.  The ones that don't ring are 
> being damped, indicating (supposedly) that the hammer is "high" under that 
> string.  Trouble is, I get the indication that all the hammers in the low 
> tenor are high on the right.  Yet when tipping the front of the action up 
> and viewing the strike points (which appear just above the flange screws 
> from this vantage point), they appear straight across.
> ( _ _ _ _ _ _ _ _ _ _ )
>    If I then file the right side of each hammer so that all 3 strings of 
> each unison are damped equally when holding the hammers against them, then 
> look at them again from the front, they appear like the teeth of a giant 
> rip saw  (|\  |\  |\  |\ |\  |\  |\  |\ , but imagine the line slanting to 
> the right --  the strike point-- slants only about a third as much as the 
> backslash.)  But when I leveled strings, nothing told me that the right 
> string of each unison was low, especially not that far low.
>    This problem doesn't occur in the section with non-angle-bored  hammers 
> nor in the capo sections.  So I'm wondering if, because of the angled 
> hammers (not in line with the shank), the shank flexes when pressing up on 
> the jack, causing the hammer to lean to the left, and giving me a false 
> indication of how the top of the hammer is shaped.
>    I've also had situations when, after spacing hammers, some of them 
> still get thrown aside slightly on a hard blow, so that one unison string 
> doesn't sound as loudly as the others.  In other words, the angle bore or 
> something else is causing the hammer to travel or lean on a hard blow.
>    --David Nereson, RPT