In a email to the reflector several months ago I reported that
the Rx340 close in dynamic range for signals closer than 10 khz
to the desired signal was only 40 db. I have done some careful
measurements on this and have found the figure is actually 65 db.
This is a much better number.
However, I decided to go ahead and see if a simple mod could
improve on this parameter by adding a tighter filter at the beginning
of the 455 khz IF chain.
At this point some review of how the Rx340 works is in order.
This is a gross simplification of how things work.
All signals are upconverted to 45.455 mhz. Here, the signals
are filtered with a 20 khz filter. Once filtered, the signal
is then converted to 455 khz. The AGC in the Rx340 is divided
up with 80 db of hardware AGC and 65 db of DSP AGC.
Up to the point of converting to 455 khz, no AGC is used
at all. All signals at this point are "full blast" inside the 20 khz
roofing filter. At this point the 455 khz signal is first amplified
with an amplifier under hardware AGC control. The added filter
must go before any hardware AGC is used.
It is then converted to 16.67 khz and digitized for DSP. The DSP
then has the job of filtering the desired signal (amoung many tasks).
Because the first local oscillator only moves in 1 khz steps to
convert to 45.455 khz, the 455 khz IF center frequency is actually
455 to 456 khz, depending on the actually tuned frequency. The DSP
performs the job of recentering over this 1 khz range to allow for
1 hertz tuning.
What I did was order a 4 khz filter module from Kiwa Electronics.
This filter has a 4 khz bandwidth at 6 db and 6 khz bandwidth at 60 db.
It centers on 455 khz. I installed this early in the 455 khz IF,
ahead of any hardware AGC control, with
reed relays to switch it in and out. The filter module is equipped with
its own amp to make it a zero db insertion loss filter.
Anyway, with this filter installed, the close in dynamic range for signals
between 3 and 10 khz away is now improved to over 100 db.
Of course, the dynamic range for signals past 10 khz was always
very very high. There is no change for signals less than 3 khz away.
That figure is still 65 db.
Because of the 1 khz center frequency tunig of the DSP, I cannot use
a tighter filter than this. The mod was not too hard to install but did
require
cutting one land trace on the second mixer board ( a little scary!).
I am now using this mod and looking for a situation where it might help.
Carl Moreschi N4PY
Franklinton, NC
|