[sdiy] VCO bleed through to other modules
Rykhaard D.A.M.I.A.N.
rykhaard at gmail.com
Fri Mar 24 06:10:45 CET 2006
In about '92, my then online eletronics mentor, gave me a couple of
tips about my power supply distribution to each module, which I've
used ever since - and I've never had crosstalk troubles. :)
First
(Sideline thought before 'First' - and I'm going to graphically make
this example and post it on my webpage, right after this message is
finished - to hopefully help others)
>From my power supply (Power One bipolar 12V @ 1.7 amps, tuned to
bipolar 15.0 @ 1.5 amps) is sent directly to a:
- positive supply bussbar
- ground bussbar
- negative supply bussbar
- Every PCB, gets it's power feed from these bussbars, through a 4
conductor 22 guage shielded cable.
- The 4 cond. cable ending at the PCB is feeding;
-- a 100uF electrolytic cap. for the +V, through a 22ohm 1/4 watt
resistor. (3 of which have saved me in the past few months, on 3
shorts. Thank you resistors!! :D )
-- a 100uF electrolytic cap. for the -V, through a 22ohm 1/4 watt resistor.
-- the ground plane, of the entire PCB
UNTIL this building start (July 2005), I'd rarely been using 0.1uF
ceramic caps as buffers on chips of my PCBs. I'd read a LOT before I
started any building last year, and have 1 x 0.1uF cerm. cap., on
EVERY single chip, as close to it's power pins as possible.
On THIS building session, I start twisting multiple wires to the pots
and / or jacks from the PCB, together, to hopefully help cancel out
any potential leakages of signals. (I had NOT done this in my
previous building, from '91 to '98.)
>From '91 to about '94, I used 1/4" jacks in my 1/8" aluminaminiumnium
( ;) ) panels, withOUT an extra ground wire to any of them.
In about '94, I switched to (then known as EF Johnson) banana jacks,
which I've used ever since. (With a banana to 1/4" conversion pair
for my outputs.)
I've NEVER, had any crosstalk. :) I've been told a couple of times,
that my 100uF caps. are overkill - but as soon as I started using
them, the previous problems that I was having with more than 2 or 3
modules, completely disappeared. :)
(From all of the reading, and recent digital work - I FINALLY learned
how well buffer caps per chip, WORKS. :O In my previous 'Deathlehem
Machine' builds, every single one of the 6 of THEM had ...... (what
would I call it?) cross-CLOCK problems. :O (Whether they were CD4011;
LF347 or 2N3904 based clocks / oscillators.)
K. Off to make a graphic example of this, for my webpage, which will
be posted tonight. :)
I hope this helps. I can't at the moment, think of any other
solutions, that haven't been already mentioned. :O :)
--
Take care,
Warmth and Peace,
Ryk
My NEW webpage home:
http://www.sdiy.org/damian/index.html - still transferring all from
the old site - Mar 1906
http://deathlehem.bravehost.com/damian.html - my modular synthesizer,
D.A.M.I.A.N.;s webpage
More information about the Synth-diy
mailing list