[sdiy] WNoise Gen Isolation

[Michael Ruberto]

Thanks for all the tips! I'll let you know how it works out later this week.
I am just finishing the protoboard for my CV scanner and I'll be testing it 
for the next few days.

M. A. Ruberto


>From: harrybissell <harrybissell at prodigy.net>
>To: Tim Daugard <daugard at sprintmail.com>
>CC: Michael Ruberto <frankentron at hotmail.com>, synth-diy at dropmix.xs4all.nl
>Subject: Re: [sdiy] WNoise Gen Isolation
>Date: Sun, 25 Sep 2005 21:11:57 -0400
>
>Sounds like good advice, Tim...
>
>I'd suggest much the same.  The actual front end of the noise source
>should be isolated with an RC network.  At this point, the impedance is
>quite high so the resistor can be larger, and the cap smaller.
>
>If you use a single opamp stage, and run the gain high enough to get
>into
>clipping you are asking for trouble.   I use two stages, and take a
>modest
>gain in each. This gives better bandwidth as well.  Maybe a gain of 10x
>followed
>by a gain of 10x.  AC couple those two stages so that you don't amplify
>the
>offset of the previous stage.
>
>You can put RC filters on the opamps... but if you don't clip them they
>should
>not generate much noise on the rails.
>
>The noise source should be kept away from sensitive inputs, and it might
>even be a good idea to use shielded cable from the noise output to the
>output jack. If that is a bananna jack, terminate the shield at the PCB
>end only.
>
>H^) harry
>
>Tim Daugard wrote:
>
> > From: "Michael Ruberto" <frankentron at hotmail.com>
> > > On my old modular I used the circuit from the Minimoog but
> > noise bleed was
> > > very bad...
> >
> > The first noise module I built had the same problem - noise
> > bleeding into everything.
> >
> > > I'm worried about the noise bleeding onto the power and ground
> > rails. What
> > > is an effective way to isolate this noise circuit from the rest
> > of the
> > > system?
> >
> > I leave the ground rails alone. I use a (semi-) star ground
> > system and figure that the ground is robust enough to handle the
> > noise. But then as I tried discussing earlier, I use current
> > paths instead of voltage paths for signals.
> >
> > For the positive rail (or both if you use a three wire supply), I
> > put a 100 ohm resistor in the module power supply path. This
> > increases the impediance of any signals trying to bleed through.
> > There are capacitors to ground at both ends of the 100 ohm
> > resistor. I use my normal bypass cap at the power supply end of
> > the resistor. At the circuit end of this resistor is lots of
> > capacitance. I use a large electrolytic as a second power supply
> > filter cap (and theorectily the actual voltage supply for the
> > module). I also use a medium and small ceramic for further noise
> > isolation (10 - 100 nF for the medium, 50 pF for the small). This
> > reduces the noise signficantly.
> >
> > The next noise module I build will also have a mucher lower
> > buffer gain. I've found that with other signals in the 10V ptp
> > range, 1V of noise is enough. I though about adding inductors in
> > to the power supply line, but have't tried that yet - inductors
> > are to hard to find to use this way.
> >
> > Remember to use the right size resistor. The first resistor I
> > used was 1/8 watt. It got hot! I bought some 100 ohm 1W resistors
> > to use in these types of power supply filters.
> >
> > The same filter design is used in my low noise - instrument
> > preamp module. The other thing to remember is that the modules
> > with these filters will be charged and operating long after the
> > other modules are powered down. These filters also require time
> > to charge. I give the noise module a 3 to 4 minute warm up before
> > making any adjustments.
> >
> > Tim Daugard
> > AG4GZ 30.4078N 86.6227W Alt: 12 feet above MSL
> > http://home.sprintmail.com/~daugard/synth.htm
> >
> > not spell checked or proff read. Don't bother telling me wruds
> > are mizspeled.
>