[sdiy] new analog sequencer

[Gur Milstein]

Hi guy's.
After almost a year of prototyping a analog sequencer I came up with the
following :

The design uses banana jack's and PCB mount toggle switch's .
The panel is design to fit in to MOTM cabinet but is also possible to be
mount in a regular 19" Rack .
The design had took me more then 8 month to reach a sequencer which have a
highly precise cv out and the right grove transition between the steps.
Trying many deferent counters and mux combination's I ended up with the
4516 and 4514 which are Obsolete chip's but there transition is the most
musical I have found  ,this are used to generate the basic gate out of the
sequencer .

The cv out has a deferent story as I wanted the sequencer to be more then
just a rhythm composer or a time event controller , I want it to be a real
tune composer in which musical tunes can be played and by that eliminating
the need for a midi-cv converter in the studio .
But this is not an easy thing to achieve since temp drift of pot's is one
problem but also there Is a need for a precise voltage reference to feed
this pots .

The common problem in many sequencer's which I played with were that you
leave the system after a long day of finding your musical tunes but the
next morning some of the steps voltage had change and now you have to try
and re adjust them , the problem is even greater when you use a quantizer
while the output of the quantizer is randomly swing between two notes.
Many of you would say well if a precise cv is needed then let's combine a
encoders pot's and a D/A and then were not only get a precise cv out but it
is also quantized .
Well I don't think so , the transition between steps is what makes analog
sequencer's deferent from other digital based sequencer , that means that
the transition between cv steps should be made by the clock of the song and
not by an internal high freq. clock that follows the analog clock.
There for I made the cv core using the LT-1236 as a precise 10V ref which
then feeds Bourns Conductive plastic or cermet pots which are multiplex by
the DG506 analog mux  which its output is blocked by the MXL1013 so there
would be no current running trough the mux which can create a R(on) that
changes the precision of the cv out.
After I try it I realized that it do  have the step's transition sound just
analog  but it was also very accurate and stable over temperature changes .

Now that's not the end of the story (yet) since In addition to achieving
this goals I have the main goal which is to clock my modular setup only by
an analog clock (vco) and eliminate the use of midi pulses or other CPU
based sequencer's as the clock of the song .
>From that point I realized that I'm gone need many sequencer's which runs
at a divide clock rates and they are controlling one of the other and by
that creating synchronized and a-synchronized time events .
The problem that occurs to me was that analog sequencer's cost a lot of
money and I have asked my self why , so I checked and realized that the
pot's ,toggle,and jack's makes this units expensive , so I decide to attack
each by a deferent method .

The pot's are the most expensive part in the sequencer and be cause I
wanted a precise cv out I need to use precise pot's and that made the $
problem even bigger .
The solution to that problem had came to my mind while I was looking in a
book which contain some musical score's .

16 pot's should be considered as 16 deferent notes which in by playing them
at deferent order can create a composition which is long and probley
sufficient for the ordinary composer and not a repeated pattern which makes
me boring very fast .
there for I created the Step Select Input (SSI)  which allow the composer
to activate each note at the right time and order he wants.
But that's not all there are 16 banana jack's for Single Step Trigger
Output and two SSI enabling input's and by patching part of this 16 outputs
from one sequencer that runs at a divide clock rate to another sequencer
SSI inputs and enable input's a sufficient note composition is happening .
The SSI enabling input and rest are allowing to activate the pre-set SSI in
a synchronized and a-sync way concerning the regular 1,2,
..16 stepping  .

The same solution concerning $ is achieved by the SSI towards the toggle
issue as there is an SSI gate output jack that sum the regular gate which
are created by the 16 toggle's position and the gate that is created by
pulsing of one or more of the 16 SSI (keep in mind the SSI can be pulses at
a divide clock rate meaning interesting pre-set grove patterns).

The one remaining issue was the jack's , which one I'm gone use .
There are 44 jack's on the panel and that is a lot and since I wanted this
sequencer to be a quality one so he could sit peacefully by the MOTM system
I realized I need to use quality jack's like the MOTM.
But then I asked my self do I really need shielded cable's to run logic
signals from the sequencer's and in between the sequencers , well its
obviously I don't as there is no problem of cross talk of audio signals.
Also there is a great need for multiple when patching between sequencer
(SSI etc) .
Banana jack's are the cheapest way and are stack able so I decide that they
would be the best choice for the control system in a modular setup.

Well that's about it concerning the design I made , I have 1100 pads to
layout on the PCB and the PCB is gone Be big and full of electronic goodies .
I might offer this sequencer as a kit to cover PCB cost but I'm not sure
yet as then I should buy lots of part's and I don't have the time and money
for that.
May be just the PCB and panel 
.I don't know but I would love to hear form
you guy's what you think ?

Thank you all
Gur Milstein

ACS  
-------------- next part --------------
A non-text attachment was scrubbed...
Name: TM-116-ALL.gif
Type: image/gif
Size: 7102 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20011018/776b856a/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: TM-116-CONTROL.gif
Type: image/gif
Size: 6549 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20011018/776b856a/attachment-0001.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: TM-116-LEFT.gif
Type: image/gif
Size: 7312 bytes
Desc: not available
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20011018/776b856a/attachment-0002.gif>
-------------- next part --------------