[sdiy] SDIY weekend, and trawl for info

[Neil Johnson]

All,

Well, here we are the Monday after the annual Synth-DIY weekend at Paul
Maddox's, a weekend filled with beer, pizza, knob-twiddling, more beer,
tea, more knob-twiddling, and a brief tour through some of the list
members' CDs (Haible, Bissell, Allgood, etc).  Thoroughly enjoyable.

Who was there?  Me, nothing much to show as I'm in ideas gathering mode;
Tony Allgood brought along his ever-expanding Oakley monosynth and his new
Orbit; Merv Thomas, from just down the road; Chris Crosskey brought along
his Formant/Digisound hybrid monosynth and intermittent keyboard; Colin
Fraser demonstrated his sequencer, running Tony's monosynth and Orbit at
various times over the weekend; and Paul Maddox and his blue room of PPGs
(Wave 2.2, 2.3, Waveterm, EVU) and Monowave.

Relevant websites:
Paul	http://www.wavesynth.com
Tony	http://www.oakleysound.com
Neil	http://www.njohnson.co.uk
Colin	http://www.octavo.demon.co.uk/colinf/home.htm
Chris	http://www.c5.cybes.org
Merv	???

Now down to business.

Does anyone around here have footprint details for the 2SC1583 dual
transistor?  Enough info to make up a PCB component symbol for layout.  A
copy of the datasheet (in PDF) would be even better.

Also, for expo trannies in general, what's the preferred dual-NPN:
	- 2SC1583	(good, cheap, not in production)
	- 2SC3381	(not as good, cheaper, still in production)
	- LM394		(v.good, expensive)
	- MAT-02	(v.good, not in production)
	- SSM2210	(as above)
	- CA3046	(real cheap, easy to do dual expos like in SEM)

And any others.

>From the above list, apart from the CA3046 and 2SCxxxx they all seem to
come in DIP8 or 6-lead can packages, so I'm planning on making my VCO PCB
accomodate 2SC1583, 2SC3381 and LM394 + clones, which should cover most
options.  Any other thoughts/opinions??

Secondly, one topic of discussion over the aforementioned weekend was the
DC behaviour of the Gilbert cell VCA.  While it is pretty good on S/N, as
it can handle several volts of signal, one problem it faces (brought up by
Tony) is due to the offset of the servo opamp, which generally being part
of a dual or quad does not have offset trim.  So, my suggestion is:  add a
trimpot to the servo opamp's +ve input to trim out its offset, and thus
avoid the click as the CV passes the offset voltage point.  Or even move
the offset slightly negative (by a couple of mV) so that the CV will never
cross the offset voltage.  Again, thoughts on this appreciated.

Tony, Paul: I've standardised on a single connector for signals/CV on my
modules now, so it should be easier to accomodate your requirements.

Neil

--
Neil Johnson :: Computer Laboratory :: University of Cambridge ::
  http://www.cl.cam.ac.uk/~nej22           +44 (0) 1223 763 646
----  IEE Cambridge Branch: http://www.iee-cambridge.org.uk  ----