Odp: Odp: [sdiy] inversion, CMOS, etc.

Roman modular at go2.pl
Tue Aug 6 19:48:31 CEST 2002


Hi
inline...

> This is what I imagine the half wave rectifier
to look like.
>
>      +5V/-5V   ---|---D|------   +5V/0V
>                   |
>                   R
>                   |
>                  GND

I'd put the resistor on the other side of the
diode, but it'll work anyway
provided next module input impedance isn't too
high.
With this you get wave shifted down by diode's
voltage drop.
With another circuit, when diode is connected by
anode to ground and
by cathode to resistor, there's no 0.6V drop, as
signal passes thru
resistor unaffected in positive halves, but there
will be 0.6V negative
pulses when the input gets clipped.
To avoid the drop, use ideal diode circuit. BTW,
the link Jim posted
about diodes is really fine reading to anyone:
http://www.oldcrows.net/~patchell/archives/idealdi
ode.html


>
>
> What I'd like to do is combine the HW rectifier
with the passive summer to
> create a pot crontolled sort of "sunrise" effect
of the wave going up and

if you add another resistor to this circuit with
diode going to ground,
you're in the business. Second resistor will add
offset voltage to the
'clipping point' of such rectifier. The more you
add, the more you
shift the wave that diode can see, causing less %
of each cycle to
be clipped. But then amplitude is attenuated by
used resistors
ratio, so I guess it's no use for you. In that
case, run additional
current from current source over the resistor.
Current source
can be voltage controlled to suit yur needs.
Additional voltage
drop caused by your current source will shift the
signal up,
causing more of it to pass thru when you want it.
I should better draw a shematic of that...


> source.  The thing I'm wondering about is if for
ultra high frequencies (RF
> type) do I need a special high speed diode for
the rectifier or will a
> 1N4148 type do the job?

If you mean real RF, not that DC rubbish under
100MHz, I'd go for
fast Shottky or PIN diodes and small resistor
values. Ideal diode
approach might not work good enough above 100kHz,
at MHz
frequencies you'd need opamps with slew rate in
range of kV/us.
It's better to use then diode/resistor rectifier
and compensate diode
drop somewhere else in the circuit.
For example another summer after rectifier.

Roman




More information about the Synth-diy mailing list