[R-390] Tuning SSB

[Jim Whartenby]

Francesco

My point was that the product detector is not coherent but the Costas Loop is as is the lowly envelope detector.  One may be able to set the product detector oscillator exactly on frequency but there is no phase control as there is in the Costas Loop aka synchronous detector.  So the product detector is not coherent.  Noise wise, there is perhaps no discernable difference between the product detector and the synchronous detector for low to moderate level signals as there is when either is compared to an envelope detector.  

It appears that there is still a question of how much of an improvement there is comparing the product detector with the envelope detector.  Since the sharp filters reduce the bandwidth in SSB, that is what is most noticeable; reduced noise by 3dB due to reduced bandwidth.  The graph shown below is again comparing the synchronous detector, not the SSB product detector with the AM envelope detector.  Not known is how low a signal that can be detected with the product detector.  It is obvious that the synchronous detector can retrieve a signal way down in the mud as is mentioned by John Costas in his 1956 paper.




The Costas Loop is more flexible in that AM, DSBSC, USB, LSB, frequency and phase modulation all can be detected.  It is so impressive in operation that every digital cell phone contains a digital embodiment of the synchronous detector.  John would be amazed!

Regards,
Jim

Logic: Method used to arrive at the wrong conclusion, with confidence.  Murphy 

    On Tuesday, October 1, 2024 at 07:03:07 AM CDT, Francesco Ledda <frledda at att.net> wrote:   

 Jim,

I designed Costas loops for coherent optical receivers, and I never came across any associated noise reduction theory associated with such loop. The only noise reduction was due to the sideband suppression that an IQ receiver can provide. 

It is not clear to me how a product detector can reduce noise. The product detector has better conversion efficiency, than a single diode detector, that may be interpreted has noise reduction.

Being an engineer, I need to see the math to understand the noise reduction process claimed in this chain of emails.

Best, Francesco K5URG


 

  
Sent from my iPad

> On Oct 1, 2024, at 03:08, Jim Whartenby via R-390 <r-390 at mailman.qth.net> wrote:
> 
> The link seems to me to be blurring terms.  The product detector is described as equal to coherent detection which I believe may be better described as John Costas' synchronous detection.  I don't see how a product detector oscillator is coherent with the transmitted carrier.  The product detector frequency may be very close to being spot on with the transmitted carrier frequency but not so with respect to phase.  
> 
> DSBSC reception using synchronous detection has advantages over SSB in the presents of jamming according to Costas and is much less complicated to transmit.  There is no need for linear power amplifiers or high Q filters.  The original question was a comparison of the product detector with the envelope detector but it seems to have quickly become, at post #3 and after, about synchronous detectors vs envelope detectors.
> 
> Jim
> Logic: Method used to arrive at the wrong conclusion, with confidence.  Murphy
> 
>    On Monday, September 30, 2024 at 10:46:03 PM CDT, Jacques Fortin <jacques.f at videotron.ca> wrote:  
> 
> Hello Giovanni,
> 
> Noise reduction of an amplitude-modulated signal by a product demodulator vs an envelope detector is a verifiable fact.
> I have found many explanations for this, but I believe that the thread from the following shortcut will provide a valid answer to your question:
> https://www.edaboard.com/threads/envelope-dectector-vs-product-modulator.262958/
> 
> For the R-390 and R-390A cases, the AGC system action time needs to be "speeded up" when SSB is received, without changing the global AGC behavior VS the AM case.
> In other words, we need to keep the same AGC characteristic for both, because the global receiver behavior and the signal meter calibration depends on it.
> Many attempts were made up to now (and by many people) to "improve" the AGC system of these receivers and the search for the best solution continues.
> Some already available solutions uses external "add-on" circuitry to achieve the necessary goals, but simultaneously implies some compromises vs other receiver features, like the use of the MGC mode and the Noise Limiter circuit when the product detector is activated.
> What we try to develop for some time is a self-contained solution with no "compromises"...
> In that respect, I still have some (more or

  
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1727800293921blob.jpg
Type: image/png
Size: 192777 bytes
Desc: not available
URL: <http://mailman.qth.net/pipermail/r-390/attachments/20241001/2ae47268/attachment.png>