[R-390] R-390 AProduct Detector

[Mark Huss]

Not sure this is such a problem as you think. Two ways to generate AGC 
is (1) RF Voltage at the Detector. (2) AF voltage after the Detector, 
often called AVC. For AM, Nr 1 is considered best, as it is not affected 
by the modulation level because it acts on the received carrier wave. 
For SSB, however, there is no received carrier wave. Instead, Nr 1 often 
will be driven by the BFO injection level. Thus Nr 2 is the preferred 
way of controlling receiver gain. You can get around this by tapping the 
RF level before the BFO injection, but older receivers not designed to 
receive SSB had no reason to do this.

With an External detector, however, things are different. If you set the 
receiver for normal AM reception, with a slow AGC time constant, the 
receiver AGC will act normally, substituting the sideband RF for the 
Carrier RF. It will be choppy and constantly varying in level as it is 
the modulation, but the Slow AGC time constant smooths this out. It 
would be the same as more modern receivers tapping the AGC voltage 
before the BFO Injection point. Attack time on the first syllable may be 
improved. And you have to have a long enough decay time to cover pauses 
between sentences. And you will have to use the receiver's filtering for 
most of the selectivity.

The beauty of the Softrock solution is that it has its own AGC, 
implemented in software, with a range of greater than 60 dB (dependent 
on the dynamic range of your sound card). Taking a worst case example of 
two closely spaced signals a hundred Hz apart, and ignoring the 0.1 kHz 
position of the Bandwidth control. Receiver AGC will be controlled by 
the stronger, unwanted signal, the wanted signal suppressed, say 30 dB 
by AGC action. By setting your external detectors bandwidth to, say 100 
Hz, and off-tuning the external detector frequency (can you say 
'Passband Tuning :-) '), the external detector's AGC will easily recover 
the suppressed signal. Or we can use an alternate technique not seen 
often any more. Set up a notch filter in software to take out the 
offending signal. Where the Receiver AGC is set by the unwanted signal, 
the Softrock AGC is set by the wanted signal, the unwanted signal not 
affecting the Softrock AGC.

This is much like the excellent Hammarlund HC-10. In Hammarlund's case, 
they literally took a Hammarlund receiver, complete with bandpass 
filters, Q-Multiplier, Detector, AVC, and Audio Amplifier and boxed it 
up. 455 kHz (430-520 kHz, set by a trimmer inductor) is taken in. 
Converted down to their 60 kHz IF whose gain is controlled by its own 
AVC, and detected it. I believe the AVC is available, but seldom used in 
practice.

I remember seeing someone who had a Drake 2B laying around unused. With 
a few wiring changes, he implemented an HC-10 clone. and Drake 2B 
receivers are going for about $180 on EBay. The HC-10 is going for +$300 
on EBay. The Softrock is by far the most expensive solution. $20 for the 
Softrock, and $700+ for the computer/Display/Sound card to use it. Of 
course, if you just happen to have a reasonably new computer sitting 
beside the R-390 that you use for logging/etc., the solution drops 
tremendously in price to about $40 in parts (connectors, box, power 
supply, cables, and kit). Plus it adds passband tuning, linear 
detection, a multitude of filter options, limited scanning capability, 
digital recording, CW and Digital Mode decoding, Propagation prediction, 
even an Atomic Clock(if it is connected to the Internet). Heck, with a 
little breadboarding and some programming, you can do away with that 
pesky Zero knob by counting the three oscillators and sending the count 
to your PC via the serial port! An eight-input mux, three 74HC4046's, 
and a PIC will do that. Tap the oscillators with #30 wire-wrap wire 
wrapped around the tube under the shield.

2002tii wrote:
> Mark wrote:
>
>   
>> A forth [SSB option] is to add a separate SSB detector/BFO at the IF
>> Out. Several Mil-Surplus detectors do this, as well as the Hammarlund
>> HC-10. The latter has the advantage of providing notch filters and
>> bandwidth filters, though it is expensive at about $300.
>>     
>
>   
>> The latter option just got a lot cheaper if you have access to a
>> reasonably fast PC. A lot of Hams are playing with a little device
>> called a SoftRock.
>>     
>
> One issue with outboard detectors is that they often don't feed an AGC
> signal back to the radio, so the AGC won't be based on the signal you
> are listening to. This is easy enough to do with the 390A, and I
> believe some of the Mil units do it. It's not necessarily a horrible
> problem, but it's not optimal, either, particularly if the outboard
> detector has its own IF filters.
>
> I have always intended to design an outboard IF section with its own IF
> filtering, selectable upper, center, and lower BFO injection, 2 or 3
> tunable notch filters, a product detector with synchronous detection
> capability, and an AGC feed. The ideal implementation would switch the
> BFO offset with the IF bandwidth, so the upper or lower BFO would
> always be just outside the selected passband. But I have always had
> other, more pressing projects....
>
> Best regards,
>
> Don
>
> _____________________________________________________________
> R-390 mailing list
> Home: http://mailman.qth.net/mailman/listinfo/r-390
> Help: http://mailman.qth.net/faq.htm
> Post: mailto:R-390 at mailman.qth.net
> Unsubscribe: http://mailman.qth.net/mailman/options/r-390
>
>   


-- 
----------------------------------------------------------------------
Some people are like a Slinky .. not really good for anything,
but you still can’t help but smile when you shove them down the stairs.
----------------------------------------------------------------------