[Elecraft] RE: PowerSDR/IF stage v0.91 Released

[Lyle Johnson]

> I think I have found a supplier for a 32.800 Mhz crystal, which would 
> put the softrock at around 8200 and make it a more reasonable 15000hz 
> offset.  I hope that this will still be far enough away from the 8215 IF 
> to not cause any problems, while making the bandpass a little more 
> symmetric.

Be _very_ careful to put a high reverse-isolation buffer between the K3 
I.F. output and any sort of mixer like the Softrock that leaks a lot of 
L.O. energy onto its input lines.  The buffer in the K3 is an impedance 
transformation device, not a reverse-isolation stage.  It assumes the 
attached device does not inject any appreciable energy back into the 
I.F. output tap, and certainly not any energy at the 8.215 MHz I.F.

A better approach when using a quadrature sampling detector (QSD, or 
Tayloe mixer) may be to mix the 8.215 MHz I.F. to a very different 
frequency, then use the Softrock at that I.F.

For example, use a high reverse-isolatin amplfier at 8.215 MHz to feed 
an H-mode mixer driven with an oscillator at 19.44 MHz.  This will 
result in an IF of 11.225 MHz where the Softrock can operate.  4x 11.225 
is 44.900 MHz.  19.44 MHz and 44.90 MHz are standard (currently in stock 
at Digikey) frequencies, and with careful construction and a reasonable 
reverse-isolation input amplifier, there is little danger of interfering 
with the K3.  I have not tried this, so use this information entirely at 
your own risk, YMMV, etc.

But, in any event, a high reverse-isolation amplifier is an absolute 
must if the device you are attaching to the K3 I.F. output has _any_ 
chance of leaking energy into the K3 anywhere near its I.F. Use good 
quality coax, keep RF out of the IF connector (the end of your 40 meter 
dipole 15 feet away from your radio, that sort of thing...).

The SDR-14 and SDR-IQ appear to be safe in this regard as the input goes 
to a buffer amp and then directly to an A/D converter clocked at about 
65 MHz.

73,

Lyle KK7P