[Lowfer] Audio Noise Floor Measurements on my TS450S

[JD]

Many receivers are in exactly the same boat Garry describes.  In my own 
case, it turns out the later specimens of the Kenwood R-5000 are often 
exquisitely sensitive and remain just as clean as necessary down to 30 kHz. 
At least, that's true of the 50 ohm input, provided that you furnish it with 
signal from a more or less true 50 ohm source.  (All bets are off with the 
500 ohm wideband transformer input, which is pretty poor at lower 
frequencies.)  There is no attenuator in line at LF unless you switch one in 
manually.  Only the 0.5-1.6 MHz range is attenuated automatically.

Since I can use SAQ-rx fairly successfully with the spare computer and its 
48 kHz sound card sampling rate, that means on an ordinary day I only have a 
6 kHz gap in coverage between 24 and 30 kHz--which, unfortunately, is where 
many of us need coverage right now.  I have one of Todd's modified AMRAD 
converters that does a splendid job but is a little less than convenient in 
the field because it needs 24 volts, and it also magnifies a certain 
shortcoming of the receiver.  I use it for special occasions like SAQ tests. 
My long term solution is probably going to be another upconverter, but this 
time only to 100 or perhaps 200 kHz!

Why not go all the way to 3, 4, or even 10 MHz?  Because the R-5000 doesn't 
have enough thermal stability.  With a synthesized receiver, the higher the 
dial frequency, the greater the absolute shift in output frequency for a 
given parts-per-million LO change.  For the temperature variations in my 
field from mid-afternoon to midnight, the receiver I most often use may 
drift 2, 2.5, or even 3 ppm.  If I'm tuned to 3.0295 MHz to copy activity at 
29.5 kHz, a mere 1 ppm drift over, say, a 2 hour span is going to throw 
things off by a little over 3 Hz...even if the upconverter OCXO stays 
exactly on frequency!

For radios that can be locked to an external reference, that's obviously not 
an issue.  They can benefit greatly from a stable upconverter.

In my own case, though, any lockable LO retrofit doesn't look practical. 
However, I could achieve stability of < 0.13 Hz under the same conditions if 
the radio were tuned to a dial frequency of 129.5 kHz instead.  Not perfect 
for the slowest modes, but a huge improvement!  The radio itself is 
reasonably spur free in the range of 100 to 140 kHz, so _if_ I could keep 
LORAN pulses from sneaking through, and _if_ I can do the frequency division 
from a 10 MHz OCXO without introducing too much jitter (and hence, phase 
noise), then I could fill my coverage gap for little by taking afvantage of 
what the receiver is already capable of doing.

At least, that's the hope.

73
John