Electronic Sustain Timer

[Carl Meyer]

This is a multi-part message in MIME format.

---------------------- multipart/alternative attachment
Sarah, you're on!  I've been in electronics all my adult life.  I also =
have a strong interest in sustain.  Let me think about your proposal a =
bit and I'll get back when I have a moment.  The first order of business =
is to define sustain.  Can't measure it if is isn't defined.  I'll post =
later about some of the problems I see over the hill.  Thanks.

Peace to you too!
Carl Meyer  Assoc. PTG
Santa Clara, California
cmpiano@attbi.com=20
  ----- Original Message -----=20
  From: Sarah Fox=20
  To: Pianotech=20
  Sent: Wednesday, November 20, 2002 10:25 AM
  Subject: Electronic Sustain Timer


  Hi all,

  I thought more about how to time sustain while taking my morning =
shower (ordinarily a very productive time for me).  It occurs to me that =
the best way to do it is electronically -- with a simple circuit made =
just for that purpose.  Many of you work at universities, and you could =
probably find a departmental electronics technician who could put =
together such a circuit on a breadboard in no more than an hour or two.  =
Some of you may know enough about electronics to do it yourselves.  As =
for myself, I once maintained an electonics test bench and =
designed/built these sorts of circuits, but I've been downsizing, and =
most of that stuff is either given away or stored in boxes.  I won't be =
able to do this in the forseeable future.

  Anyway, here's a description of the circuit.  This is very elementary =
stuff, and any electronics tech should know how to put these circuits =
together out of no more than a few dollars in parts (not counting the =
breadboard, which you would borrow):

  There would be two op-amp circuits with 100-fold (40 dB) differences =
in gain.  Inputs to both would be capacitively coupled from the same =
microphone.  (An adjustable-gain preamp circuit would be nice here.)  =
Outputs from the two amps would be capacitively coupled to two unity =
gain op amps ("voltage followers."  Outputs from the voltage followers =
would then be quarter-wave rectified and low-pass filtered, so as to =
yield an approximately DC output.  (i.e. diode, serial with resistor, =
feeding to one lead on a capacitor, with the other lead grounded).  =
Although this voltage wouldn't be RMS, it would be "similar" and would =
summate energy not only only from the fundamental but also from the =
harmonics.  Those who want true RMS can possibly find specialized V^2 =
amplifiers, but that's a bit expensive/weird/exotic.  Feed the outputs =
of those circuits to two comparator circuits that would detect a voltage =
threshold.  (Assuming the 100:1 gain is accurate, the threshold voltage =
should be the same for both the high- and low-gain circuits.)  Finally, =
feed the outputs of the comparators to logic circuit that would output H =
when the high-gain circuit's output is H and the low-gain circuit's =
output is L.  This could be done with an XOR chip, I think.  Use this =
output to trigger a millisecond timer (e.g. an old Hunter, probably =
available for loan from the junk pile of any dept of psychology).

  How to use it:

  (1) Hook a soundboard mic to the circuit's input.  To avoid =
constructive/destructive interference effects, mute two of the three =
unisons of the note being tested, both in the main *and* duplex =
sections.

  (2) Play a note loudly enough to exceed the low-gain circuit's =
threshold (loudly enough that the timer won't run immediately.

  (3) As the amplitude falls, the output signal of the low-gain circuit =
will fall below threshold, and the timer will start running.  The sound =
may "warble" a bit over the threshold, resulting in the timer starting, =
stopping, starting, stopping, and then running.  That's OK, because the =
same thing will happen when the timer shuts off.

  (4) As the amplitude falls further, the output  signal of the =
high-gain circuit will also fall below threshold, cutting off the timer =
-- again probably with some warbling, which is OK.

  (5) The 40 dB decay time can then be read in ms.  Assuming a linear =
falloff (and I don't know if this is a valid assumption), the 6 dB =
falloff (i.e. amplitude half-time) would the 40 dB decay time divided by =
6.67.  If decay is nonlinear, then just report the 40 dB decay and try =
to reference it to some initial SPL value.  With regard to SPL, 99% of =
the world has no clue that an SPL reading is utterly useless unless it =
is referenced to a distance from the sound source.  Owing to the "large" =
sound producing area of the piano and the complex nature of its =
acoustics, I would probably put the meter on a soft blanket on the floor =
beneath the midpoint of the "strip" of soundboard immediately under the =
string, with the mic pointed straight up and the piano lid closed.  =
That's probably as reproduceable as you're going to get.

  For those who aren't knowledgeable in electronics, don't worry.  This =
isn't neary as comlicated as it sounds.  It's a very easy project for =
some electronics technician.  Look in the following departments: =
physics, any engineering, neurobiology, psychology, RTF, possibly even =
music (???).

  Peace,
  Sarah

---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/pianotech.php/attachments/52/43/93/41/attachment.htm

---------------------- multipart/alternative attachment--