>     nothing of more easy. I connected a signal generator to a QSD
driving
> the sound card. Then I started Winrad,
> and captured the following two pictures. The first :
> http://www.i2phd.com/images/image.gif
> is what what you see when the unbalancing is total, here the image
> response is even stronger than the true signal.
> The second :
> http://www.i2phd.com/images/noimage.gif
> is what you see when the balancing is done at its best. The image
> response cannot be seen anymore. Of course you
> cannot maintain such a good rejection all over the band, for this you
> need an automatic algorithm, not a manual
> balancing. From my experience, sampling at 96 kHz, if you do an almost
> perfect balancing in the middle of the upper
> (or the lower) half, you can count on approx. 45 - 50 dB of rejection at
> the band extremes.
> This at least with my hardware.
> 
> 73  Alberto  I2PHD 

 Alberto,

 Thank you really much fellow!!! So nice that you went through all the
trouble to make those print-screens.

 Ok, they look exactly what I pictured in my mind. Well, I've done now
trick or two and can have about 30db image rejetion - which definently
is not enough. Something is wrong, because any software I've tried
(powersdr and both of your softwares) don't have enough range for the
phase adjustment. The Powersdr has less range than the others, giving
of course worst image rejection. There are now two possible reasons
for my problem (I already remade the cables): It's either the
soundcard, or then my IQ local oscillator is not giving pure IQ, it
can be few degrees off, i have to examine that!

 Alberto, can you say how many degrees does Winrad give max ? It would
be nice to have some kind of idea how far we are.

Thank you and 73!

-Janne, OH1GTF

Reply via email to