Hello Andre,

I believe the SDR bandwidth is dependent on antenna impedance. Also,
depending on how you have arranged your op-amps, the gain may be
dependent on antenna impedance as well. You can remove the gain vs.
antenna effect by using differential op-amps (also known as
instrumentation amplifiers).

All this is explained in one of the four Youngblood papers on SDR:

Youngblood's "SDR for the Masses" 4-part QEX article.

http://www.arrl.org/tis/info/sdr.html

Of course, using a pre-amplifier in-front of the QSD removes these
affects. You can also try an antenna impedance matching circuit. I'm
not sure whether a high-impedance input pre-amp would work, at least
not without an isolator or circulator to mitigate mismatch effects.
Speaking of which...

There is an interesting type of op-amp isolator or circulator circuit
that was originally realized by Wenzel. You might try putting one of
these Wenzel circuits in-front of the QSD, I don't know anybody that
has tried it yet; be sure to select low-noise amplifiers. The Wenzel
circuit uses voltage amps, current amps typically go higher in
frequency. But these days finding suitable voltage op-amps that can
cover 1-30MHz should be easy. Do a Google search for "isolator
circulator Wenzel" I don't have the link at-hand.

What ever you do, don't forget to take into account your input
band-pass or low pass filtering.

Good Luck, David

--- In [email protected], FRANCIS CARCIA <[EMAIL PROTECTED]> wrote:
>
> a 6 dB resistive pad works for me connecting to the IF of my RX
> 
> Andre Belokon <[EMAIL PROTECTED]> wrote:          Hello
> 
> I recently studied schematic of SDRZero
> http://py2wm.qsl.br/SDR/SDRZero-2.html. this sdr receiver include rf
> amplifier (MMIC variation) between bandpass filter and QSD. author
> wrote: "An RF amplifier follows, the main reason to use one here is to
> present resistive impedance to the QSD (quadrature sampling detector)."
> As I can understand connecting QSD to source with reactive impedance
> will produce phase/gain imbalance. is this correct? someone research
this?
> 
> Regards, Andre
>


Reply via email to