synchronous am detector

[km6sn]

Apparently synchronous AM detectors have substantially improved immunity to selective fading.

Reading on the subject illustrates an implementation method therefor; namely, sampling a bit of the receiver IF and feeding it into a limiting amplifier. The output of the limiting amplifier is used as the LO for a mixer. The mixer input is IF, and the mixer output is audio which must be low-pass filtered.

I believe this is an elegant solution, in that it will track a drifting station, and accommodate differing frequencies of stations in a round-table. It also provides for continuously variable selectivity: just change the rolloff frequency of the audio LPF.

Has anyone used this technique? What were the results? I can't wait to implement and test it, but I have two projects ahead of it.

73, Rod

[Tom WA3KLR]

Rod,

My Drake R-7 receiver uses that method for driving the LO port of the product detector.  The other method is to use a phase-locked loop (PLL) instead of the limiter strip.  Each method has its advantages and disadvantages.  My Sony 2010 portable uses the PLL method.  The PLL LO may hold through a deep fade better than the limiter LO, but depends on implementation subtleties like loop bandwidth, acquisition bandwidth, lock-in time, hold time chosen.  Many of the original applications for the synchronous detector were for continuous reception of one signal in point-to-point communications, so in this case the PLL acquisition annoyances are not an issue.

For ham QSOs where each guy may be on slightly different frequencies, there is a beat note until the loop pulls in, or maybe it doesn’t. The limiter LO will work instantaneously like the diode detector. I prefer the limiter method also since I am mostly listening to ham QSO's.

The main thing is that the synchronous detector probably utilizes a double-balanced modulator i.c. for the product detector and you get a very linear detection versus the diode envelope detector which will have higher distortion. However, most of my AM listening lately is on my Kenwood TS-430 which is a diode detector, the east coast AM’er sound great.

All of the above’s desired gain in performance depends on the design subtleties and the signal conditions in the bandpass at the time.  In other words, if one can switch between the synchronous detector and a diode detector there are times when the synchronous detector is no better, in my opinion, but the limiter synchronous detector should be no worse.

[DMOD]

Here is summary of SD:

https://www.electronics-notes.com/articles/radio/modulation/am-synchronous-demodulation-detection-detector.php

and a buildout article from QST of July 1993:

http://www.arrl.org/files/file/Technology/tis/info/pdf/9307028.pdf


Phil

[Tom WA3KLR]

I was getting 'Page Not Found". This is the corrected URL for the electronics-notes website (I was not familiar with them, thanks!)

https://www.electronics-notes.com/articles/radio/modulation/am-synchronous-demodulation-detection-detector.php

The filter method, having a narrow bandwidth filter, probably a CW bandwidth, has a large group delay change with a small frequency change.  This is the problem as you need to maintain the proper phase relationship between the signal and LO ports.

[DMOD]

Thanks Tom.

I fixed the electronics link. For some reason the URL got concantenated when I copied it the first time.


Phil

[km6sn]

Tom and Phil,

Thanks for the info and ideas.

I am planning on using the NE604 for the limiter, and then doing whatever I need to drive a diode DBM.

I, too, like the instantaneous shift.

I was thinking about using an good quality AM crystal filter at 6.7 MHz in front of the NE604, hoping that the wider filter would not introduce substantial phase distortion.

What are your thoughts on that?

Regards,
Rod

[Tom WA3KLR]

I don't think you want to use/drive a double-balanced diode mixer but a double-balanced modulator i.c. (does work as a mixer)  Perhaps my mistake in using the term mixer rather than modulator above.  Unless you are doing something at VHF, no need to come up with high driving power for the diode assembly.   The classic old part for use at lower i.f. frequencies was the MC1496.

[km6sn]

Hi Tom,
What do you see as the disadvantages of using a diode DBM? Increased drive level potentially getting into the IF chain?

And, here's a mind-blower: I have considered using the output of the limiter mixed with the receiver LO  to create a signal
exactly at the incoming AM signal's carrier frequency. Such a signal could be the LO for a separatee direct conversion receiver, with
amazing audio quality! (Yes, extreme shielding and decoupling would be a must!)

Please ignore the preceding paragraph as the mad musings of an OT who loves to solder!

Looking forward to your thoughts on the first paragraph, though.

Regards,
Rod

[N8ETQ]

Yo'

        For some more info on this topic you may want to
look at the CV-157/URR Manual.   The CV-157 was the
Synch detecter for the R-390 series Rx's.   It also was used
for SSB detection.    The motor driven LO is phase locked to
RX's IF and therefore the XMTR.     Yes it's slow but when
using your locally generated carrier and locked up to a
fairly solid SW station the output was amazing.    I sure
miss Radio Merlin.... 

        Not so hot for Ham QSO's  due to the slowness of the
AFC motor.   Selective fading generates a "Stereo Effect"
where the fade will "Pan" from the LSB to the USB output
or visa-versa.


http://www.militaryradio.com/manuals/CV-157/cv-157-tm11-266.pdf

73
/Dan
N8ETQ/CLE

[Tom WA3KLR]

Rod, driving a diode DBM is just unnecessary, everything can be done at low levels.

Basically, the NE604 has the resources to do the job, the limiter and the product detector which gives the synchronous detection of the AM signal.  I see that i.c. is rated up to 21 MHz.  The data sheet example and QST article above run at 455 kHz.  

I haven’t digested the QST article’s full design, but it looks like the guy kind of went overboard.  But the good thing is, he can easily switch between 2 AM detectors – Synchronous with PLL drive or limiter drive.  He also has a BFO mode for SSB detection which is what you get with the one port now driven by an independent free-running oscillator. Probably easier than doing all of these mode variations with just the NE604 and other circuit switching.

Semantics is the thing here unfortunately.  In one post above you read the theory that there are 3 ways to come up with the LO signal for the one port of the detector. Some people including the author of that article only call it synchronous detection if you are using the PLL drive technique.  He calls using the limiter drive technique as quasi-synchronous detection.  

Both detection modes use a product detector implemented by using a double-balanced modulator circuit which is made of 2 differential amplifiers (paralled inputs, cross-coupled outputs) also may be referred to as the Gilbert cell, a very similar common i.c. circuit.  This is what is in the circle with the X inside, called the Quadrature Detector in the NE604 data sheet.  It’s called that because they mainly intend the part to be used for FM detection.  When the incoming FM signal is phase-shifted by 90 degrees and fed to the other port you get FM detection.  In AM synchronous detection the product detector is operating in phase with the AM rf signal and another way to describe it really, is active full-wave rectification of the AM signal.  

The NE604 allows you to get at the ports of the product detector whereas other i.c.s such as the MC3357 with the same resources, are set up with the limiter and product detector for FM application only.  (By the way, some simple double-balanced modulator ic’s with oscillators were popular in QRP transceivers with direct conversion receive were the NE602 and NE612, which are also used in the QST design.)

Rod, you described bringing the limiter/LO signal out and doing an external direct-conversion.  This is unnecessary.  The AM synchronous detector described above IS direct conversion of the AM signal!

If you want to play around and compare all the detectors, build the QST article. Otherwise, the NE604 alone will do most of the job for a limiter-driven AM synchronous detector. If you do plan to build it, look for later feedback articles/mods.

[km6sn]

Tom,
Thanks for the ideas.

My only complaint about the NE/SA612 is the abysmal IMD performance. That is why I am
interested in using a DBM.

The product detector and subsequent audio filtering will determine off-channel interference,
and IMD. I plan on using little or no selectivity prior to the synchronous detector, not wanting
random phase shifts. I want the pure audio.

I have two projects in front of this, but I will be building one. I am eager to listen to it.

Regards,
Rod

[WBear2GCR]

Please keep us posted and up to date on ur design!!


                  _-_-bear

[WU2D]

Yeah keep the LO locked or at least stable; but be able to slew to each new signal on the turn overs. Big difference in slow SW broadcast vs ham. Old Buzzards excepted!

Add Antenna Diversity with simple amplitude voting, Antenna and Receiver diversity or if you really want to get nuts - Rake Phase combining...

[km6sn]

Good ideas. What is rake phase combining, and what is it for??

[Steve - K4HX]

Rake is a mathematical process of combining multiple antenna inputs or multipath components with weighted correlators to obtain the best signal fidelity or bit error rate. Most often used in digital transmission systems to combat multipath fading and signal distortion. Common in some CDMA cellular systems.