Touch Weight

[James Ellis]

During the various posts about touch-weight and related subjects, some (I
forgot who) said touch-weight was not a static measurement, because the
hammer is moving when the measurement is made.

Despite the fact that the hammer may be slowly moving when the measurement
is made, it IS a static measurement.  It is not measuring acceleration or
deceleration.  Rather, it is simply measuring how much force it takes to
overcome the friction enough for the hammer to barely move, whether up or
down.

Bumping the table simply shakes the assembly and overcomes the "sticking",
or "starting" - "stick-slip" - friction of the action parts.  The harder
you bump the table, the sooner the hammer will move.  I have done many,
many of these tests, and I find that I get more consistent results if I do
NOT bump the table.  Also, if there is a friction problem - if something is
momentarily hanging the slightest bit at one part of the stroke, it will
show up much more quickly if I don't bump the table.  I can then check to
see what part is momentarily hanging at some point in the stroke.

However, before I take any readings at all, I "exercise" the action - i.e.,
I run my finger up and down all the keys several times with just enough
force to cause the hammers to rise, and then fall back.  I do this several
times.  You will find that if the action has been sitting for some time,
the very first measurement you make will be different from the next, and
the next, etc.  After several movements of the key, the readings will all
be nearly the same.  When the action has been sitting for a while, the
parts seem to take a very slight "set".  This is not to be confused with
verdigris, sticking, or anything like that.  Perfectly new parts will do
the same thing.

Jim Ellis