[Lowfer] RX-320

[Steve Dove]

<<

Just curious if any Ten-Tec RX 320 users are using it with QRSS and what
their results are with frequency stability.  I fired mine up and even at 10
second dots the stability stinks.  It's a neat little rig and it does wahyt
it's designed to do but it would be nice as a back up on QRSS. 
Is it just mine and if not, has anyone done anything to settle it down? 

Thanks and Happy New Year,

Jon
W1JHJ

>>

Happy New Year to you too, Jon,  hope you are well.

The RX-320 is a better receiver at its price than we have a right to expect.

Your question about stability raises a good point.  Although no worse
than many radios costing far more and, as you mention, more than adequately
stable for its intended purpose, it really isn't good enough for long-term
or unattended reception as is now common at LF.  By this I mean using it as a 
front-end for Argo to take snapshots of a 3Hz slice of spectrum all night.  
(Unless of course you aren't too choosey about exactly which 3Hz. . . )   
It  _will_  wave about with temperature, taking some if not all of your desired 
spectral 'slice' off the screen with it.  
For short-term, manned, operation especially where the receive
frequency is regularly checked against a reference (easy by bopping over
to the other 'VFO' or a memory) it is almost livable.

The main reason for this is the three crystal oscillators in the design that
directly affect the received frequency:  Y1, the 14-odd MHz (tweakable) rock
from which the 1st. and 2nd. LOs are synthesiser derived,  Y2, the 4.6-ish 
MHz 'BFO' crystal (divided down and used to convert the 455kHz IF to the 
AF third-IF). and lastly Y3, the dsp/convertor sample rate-clock oscillator. 

Now, one crystal I don't mind 'nuking' (translated:  stabilising by Positive
Temperature Coefficient resistor, as per the esteemed Mr. Koehler),
but three definitely falls into my 'too much like real work' basket;  in all
likelyhood (based on experience with other radios) it still wouldn't achieve 
the stability of a single TCXO'd referenced radio.  It would probably be 
adequate, though.  The Loran lines would be a bit wavy overnight.

There was one other factor which stopped me making the effort;  the
AGC attack time is such that any spike or static burst forces exagerrated
AGC action, punching 'holes' which receivers with longer AGC attack
times just allow to blow through and saturate to far less ill-effect.  At
least with the (3 or 4) control packages I've tried, there is control over
AGC release time, but not attack.  Perhaps someone knows of a package
which does provide a handle on this separately  -  I must confess to not
having dug too deeply into the control protocol to see if a coefficient is
even accessible for this.  Possible.  Again, this is something which is quite 
acceptable for the receiver's intended purpose, but can ruin a good plot's
day.

Ah.  One other thing.  The exagerrated and early LF rolloff that 'stock'
units have and is addressed by Frank K0BRA's now famous transformer
transplant, was deliberate.  On the front end there is a shunt network
that throws a big notch in at about 140kHz(!) in conjunction with a
couple of series Cs (C106, 7) that are really significant at LF.
Frank's mod, plus disabling the notch and bypassing the small Cs with 
a 0.1uF results in an embarassment of sensitivity at LF.  
I'm sure that it was all there to reduce the sensitivity up as far as the 
broadcast band to hopefully allow an otherwise broad-band front-end, and
as anyone who has used one of these can attest, it pretty much works.
Modded as described, though, one will need an input attenuator and/or
filtering for LF;  this is the sort of thing 'Altoids' tins are really for. 

Don't misunderstand me.  I think the RX-320 is a great little box, but a
bit shy in a few regards as a good LF receiver.

        All the best,

                Steve        W3EEE