[Re: [Pulse wave with no DC]/4066]

[Harry Bissell]

Harry Bissell writes:
Gene,

The method is to run the 4066 (4051) between -.7VDC and +15vdc. This allows
the input to conduct slightly negative, not possible with Vss at ground.
Granted, the available signal swing is very small, but the input diodes wont
clamp until -1.4 (or so) volts. The CMOS noise immunity is decreased because
the drive signals dont go all the way to Vss (only to gnd). But with the
typical threshold of 1/2Vcc for logic switching, this is really no problem. 
   Maybe I wasn't real clear, the idea is to make a negative Vss supply by
using a resistor (to -15) and a forward biased diode to ground. Like a shunt
regulator with a .7 volt "zener". This does work, and it's really quite
elegant. I had to scratch my head when I saw it for the first time.
Harry


Gene Zumchak <zumchak at cerg.com> wrote:
Harry,

    I'm not sure what you mean by biasing the ground pin at -.7 volt.  The
inputs are +/- some
number of volts.  Just put a pair of parallel diodes across the input.  As you
say, when the
switch is on, its on resistance is low compared with the diodes.  When the
switch is off, the
diodes conduct and prevent the transmission gate inputs from seeing more than
+/- .7 volts.

Gene


Harry Bissell wrote:

> Sure, you can use 4051/4066 if you locate them at the summing junction of an
> op-amp (after the input resistors), and bias the Ground pin at -0.7V. If the
> switch is on, the voltage across the switch is set by the ratio of the on
> resistance and the input resistor, and if it is off the input resistor
limits
> the current to the input pin (diodes conduct) who cares.
> The Prophet V did it this way.
>
> harrybissell at netscape.net
>
> jorgen.bergfors at idg.se wrote:
> Hi synthesists.
>
> Yesterday evening I tested a little circuit that cancels the DC component
that
> you get when you modulate the pulse width on a VCO. Roman Sowa came up with
> the excellent idea to just add the CV to the output.
>
> I added an extra op-amp after the comparator that generates the pulse wave.
A
> 62k resistor goes from the comparator output to the inverting input. A 22k
> feedback resistor from output to inveriing input
> A 22k resistor from non-inverting input to ground.
> A 22k resistor from PW CV to inverting input. A 1k resistor from output to
> output jack, as usual.
>
> Now the wave will be automatically shifted to cancel the DC. You also have
the
> added bonus that the signal level vill be reduced to 10V p-p without passive
> voltage divider.
>
> A gotcha with this circuit is that you can't use 4066/4051 for waveform
> selection anymore. This is because the peaks now can go between -10 and + 10
> volts.
>
> Happy soldering
> /Jorgen
>
> ____________________________________________________________________
> Get your own FREE, personal Netscape WebMail account today at
http://webmail.netscape.com.


____________________________________________________________________
Get your own FREE, personal Netscape WebMail account today at http://webmail.netscape.com.