en lien avec le sujet concernant la panne de hs.dorian sur son PE-2000, et la demande de gouji concernant la modification de temps de release, je cite ci-dessous une réponse sur analogue-heaven concernant cette modification et également la modification/atténuation de l'étrange fonctionnalité de coupure des notes en "release" à l'appui d'une nouvelle touche, dans un but (semble-t-il) de gestion de la polyphonie sur ce synthé.
Il s'agit donc de remplacer 48 résistors (1 par touche, 350k à la place de 100k) pour l'allongement du release (environ 4s au lieu de 1,8s max d'orgine), et également 48 résistors (1 par touche, 100k à la place des 10k) pour l'atténuation de l'effet de coupure sèche des notes en phase de release à l'appui d'une nouvelle touche.
J'aurais été bien inspiré d'y repenser lorsque j'avais plus de temps libre il y a encore quelques mois... je garde ça en tête pour le prochain conf....
Keith R. Murray a écrit :
This post is definately for those who are comfortable with a soldering iron.
In summary, the following mods will:
1. Extend the release time (called sustain on the PE) to about 4 s.
2. Remove the annoying 'feature' that cuts off the release of notes
after key-up, when you play more notes.
VCA/ENV Board Description
-------------------------
Under the keyboard there are 4 PCBs. Each one contains 12 VCAs, 12 AR
envelope generators, and some other summing circuits. All of the
circuits are discrete and are arranged in 12 identical columns (one per
key). Each PCB corresponds to one octave (starting at F).
Starting at the top of each column, there are roughly 7 'rows' of
resistors. They are:
row 1, 6.8k 10k
row 2, 33k
row 3, 100k
row 4, 12k
row 5, 680k, 3.3M, 100k
row 6, 1M
row 7, 100k
Max Release Time Mod
--------------------
The 100k resistor in row 3 affects the time constant of the release.
The un-modded release time is about 1.8 s. If you increase the value of
this resistor, it will increase the maximum release time.
There is a tradeoff with increased release time, and how useful the
range of the sustain knob is. If the release time is extended too much,
all of the useful range of the sustain knob is squished into only about
1/8 of a turn. I found that a replacement value of 350k gives a good
max release time (4 s) without messing up the 'feel' of the sustain
knob.
There is one resistor to replace for every key (48 of them).
Release Cutoff Mod
------------------
On the PE, if you have some notes still in release (after a key-up) and
you play some more notes, the original notes will be cut off. According
to Jason, they actually call this a 'feature' in the owner's manual.
This is really weird, and at first I though the design must have a bug
they were calling a feature. But on closer inspection, I found out they
actually designed it that way...
A CV for the release time of every key is tied together on a single bus.
As more and more keys are pressed, the CV on this bus is increased, and
the release time of ALL keys is decreased, irrespective of the position
of the sustain knob. I think of it as a weird analog way of controlling
the polyphony. With an un-modded PE, the key-down of only a single key
will cause all notes still in release to cut off.
The patch point to this common bus is the 10k resistor in row 1. If you
follow the trace, and then the jumper wires, you will see that every one
of them meet at a common point on another PCB near the power supply.
My solution was to increase this 10k resistor, so that the key-downs
wouldn't add as much CV to the bus, thus limiting the effect on the
release of other notes. Just like the release mod, there is a tradeoff
between a good value of this resistor and the playability of the PE. I
found a value of about 100k works well. Note that this doesn't
eliminate the cutoff problem completely, it just takes much longer for
the notes to cutoff, to the point where you can't notice it.
Just like the release mod, there is one resistor to replace for every
key (48 of them).
-Keith
--
Keith R. Murray
AKA The Stereo Effect Project
le lien de la page sur analogue-heaven :
http://analogue-heaven.1065350.n5.nabbl ... .html#none
pour qu'il retienne looonnngtemps le son mourant grâce à ses multitudes de réservoirs en papier qui vont emprisonner les petits électrons pour qu'ils galèrent à retrouver la sortie... 
...En attendant de me pencher sur le sujet, j'ai au moins trouvé le Service Manual de la bestiole : http://www.synfo.nl/servicemanuals/Korg ... MATICS.pdf 
