On Thu, 1 Apr 2010, Geoffrey S. Mendelson wrote:

Bob Dobbs wrote:
When there is audio (modulation) present, there most certainly is a carrier,
otherwise it's suppressed and therefore problematic for sync-det.
If someone were to modulate their SSB signal with anything close to a steady
tone the sync-det could possibly get a lock.
note* - there isn't a way to engage the sync-det in either of the SSB modes on
the only radio I have that has it.


No. Most ham rigs made since 1980 don't actually produce an AM signal, they
produce a double sideband reduced carrier signal. Ham rigs produce a signal
by taking an AM signal and running it through a filter to remove
the carrier and the other sideband.

Their "AM" mode signal is made by recombining the the upper and lower sideband
signals, with only a tiny residual carrier.

Most AM receivers can receive this signal, but there is no carrier to lock
on to, so I doubt that a sync detector can lock onto them.

But that's exactly the situation for a synchronous detector.

An AM signal reaches your receiver, but the carrier is too weak due to
selective fading. The issue is the strength of the carrier to the
sidebands, if the carrier is too weak it will sound like overmodulation
(which it is, just at the receiver rather than transmitter).

You can reinsert the carrier at the receiver, but with DSB of some sort,
you don't know where to set it properly. It's not just the issue of
single sideband where a mistuned receiver will shift the audio by a
certain amount (which one can live with), but with DSB if each sideband
doesn't translate down to the same audio frequency (which would happen
if the BFO isn't in the right spot), then the two sidebands clash. If
the incoming carrier is weak, the issue of beating against the local
carrier from the BFO is a secondary issue.

A sync detector figures out where to place the missing carrier whether
the carrier is weak or not, or even whether there is a carrier. It
doesn't matter if the carrier was weak compared to the sidebands when
the signal left the transmitter, for if the carrier was weakened along the
way to your receiver.

Which brings up another issue. Everyone talks about sync detectors, but
there are a variety of schemes, which do end up operating differently.
People think the sync detector is vital, yet then "some work better than
others". Some of that may be due to the specific scheme used to provide
the feature, while in other cases it can be due to bad implementation of
a specific scheme.

The stereotype of the phase locked loop locking to the carrier of the
incoming signal and then that is used as the BFO to translate the signal
down to audio, that may be the least used. It's certainly similar to
schemes where the carrier is simply amplified a lot in reference to the
sidebands (like the scheme to use a Q-multiplier to raise the level of
the incoming carrier compared to the sidebands) or a very narrow IF strip
is used to isolate and amplify the incoming carrier and then fed to the
product detector. They work, but if the carrier disappears, they go out
of lock, and they certainly aren't as versatile as some others. (I gather
the Drake R7 used a detector along these lines).

There was a whole wave of detectors for DSB in "Ham Radio" magazine in
the seventies, and the next scheme up was to get the location of the
missing carrier by looking at the sidebands. Double the frequency of
the incoming signal, and then divide it back, and you'd get a signal
right in the middle. Since the sidebands are duplicate of each other
Use a town at the transmitter to make it simple, a 1KHz tone. A 10MHz DSB
signal will then have an output at 9.999MHz and 10.001MHz, if you add
those up you will get 10MHz, and it will remain 10MHz no matter where that
tone goes. The 10MHz is exactly in the right spot to properly translate
the sidebands back to audio, and that's what you want.

When Webb described the "synchronous detector" in CQ about 1958, the
intent was to demodulate DSB with no carrier. (The carrier doesn't carry
content, it just means you don't need a BFO at the receiver or a means to
know where it should be set, but the carrier does use power, while the two
sidebands provide information about where the carrier should be, and allow
for some level of diversity reception by selecting one of the two
sidebands.). It doesn't lock to the carrier, since no carrier was
expected. It uses information from outputs of the product detectors (yes
two of them) to show where to place the reinserted carrier. It works with
any DSB signal, whether it has a carrier or not. It won't demodulate SSB
since there is no information on where to place the BFO in reference to
the sideband, you need to unlock the loop since the PLL will otherwise
try to lock to what it can't find and thus not tune properly. This type
of detector is very similar to the "sideband slicer" and other such
products that used the phasing method of sideband reception for SSB,
allowing one to select upper or lower sideband by the proper combination
of phased signals, the addition being the circuitry to lock the BFO
to the proper place.

I gather that's the common type of sync detector in most receivers
nowadays, but I don't know for sure. One can have that sort of
arrangement, that does lock to the sidebands, without fully decoding
the two sidebands separately, and there seems to be receivers that don't
provide for that selectible sideband.

Michael