[ARC5] CW/MCW

[Fuqua, Bill L]

   Well, the puzzling part is that they knew very well those days that a mixer and product detector were the same thing.
The only difference is the output frequency range. Even a diode detector worked well as a product detector just as long
as the BFO injection voltage was much greater than the signal to be detected. Just as a single triode detector. In both cases
you are relying on the non-linearity of the device to achieve frequency mixing. If the device is driven so hard that it works
as a switch, more or less, you have a reasonably good product detector. 
   Other good options were to use pentagrid converters for product detectors. 
Some Hammarlunds use the last IF (pentode) as a product detector. 

73
Bill wa4lav


________________________________________
From: Kenneth G. Gordon [kgordon2006 at frontier.com]
Sent: Monday, November 11, 2013 3:05 PM
To: Fuqua, Bill L
Cc: ARC-5 List
Subject: Re: [ARC5] CW/MCW

On 11 Nov 2013 at 18:27, Fuqua, Bill L wrote:

> It really had not occurred to me before but one possible reason is
> that since radios of the day did not have product detectors, receiving
> CW required turning off the AVC, turning up the volume control and
> constantly changing the RF/IF gain for a suitable audio level.

Yes. That is one reason. Others here have suggested other reasons for
MCW, one of the biggest being that DFing is much easier. Try DFing a CW
signal sometime: it is very difficult. And automatic DFing is almost impossible
under those conditions.

> Without AVC you could not just leave the radio unattended monitoring
> for a CW signal, especially for a beacon.

Again, yes.

> I just don't understand why they did not begin using a product
> detector earlier except that possibly they were concerned that the BFO
> would leak into the earlier stages of IF and desense the receiver.

Well, no, not exactly. Injecting the BFO signal into the IF chain AFTER the
AVC diode was common practice in some receivers (most notably,
Hammarlunds).

Even in the later model AN/ARC-5 receivers, this was common. Witness the
12SF7 as the 2nd IF tube in the communications receivers of that block. The
12SF7 contains a diode for that specific purpose: AVC not effected by the
BFO. The AVC in those receivers is ALWAYS active, and works with the
BFO on or off.

The problem with using AVC with the BFO on in MOST receivers isn't that
the BFO itself desenses the IF chain: it is that the AVC diode "perceives" the
BFO signal as just another very strong signal, rectifies that, and thus the
resulting AVC voltage is much, much higher than that which would result
from only the wanted signal, and THIS higher AVC voltage is what
"desenses" the IF (and RF too). It simply reduces their gain as it normally
would with a strong carrier present in AM...or in MCW which is simply
another implementation of AM.

> I
> have a HQ129X that I got years ago. I have several. Actually my first
> Novice station had one. Anyway this one had an additional switch added
> to it for a solid state (dual gate mosfet) product detector. It was
> tied to the original BFO and works great as does the AGC. No leakage
> problems at all.

Yes.

One added factor concerning "product detectors"  is that the PDs of the early
periods were very much more complex than the simple diode detectors
which were most common for AM detection in most receivers at the time.

Most of those early PDs required multiple tubes to implement. Like three
triodes. Although I am not certain of this, it seems to me that Collins
pioneered a MUCH simpler "product detector" using a single triode with their
KWM-1.

Although this detector is not a true product detector (which is defined as one
in which all audio output disappears if either the IF input or the BFO input is
missing or removed) they work extremely well.

Heathkit used an enhanced version of this "Collins" single-triode PD in all
their SSB rigs.

Although I have not examined it very closely from a circuit-analysis
perspective, I have always suspected that that single-triode PD is actually an
implementation of the so-called "infinite impedance" detector.

It does demodulate AM very well too when the BFO is turned off.

I might add that all a good SSB or CW detector is, is a mixer with its output
at audio frequencies. Therefore, any really GOOD mixer circuit should work
well in that service.

Ken W7EKB