[AMRadio] AM vs. broadband line noise

D. Chester k4kyv at charter.net
Sun Sep 16 17:13:47 EDT 2007

The problem is the detection process.  The solution would be a synchronous 

With the conventional envelope detector, everything within the passband of 
the receiver beats with everything else.  The AM carrier beats not only with 
the sidebands, but with every bit of noise.  Noise components on every 
freqency beat with all other noise components within the passband.  And 
every noise component beats separately with the upper sideband and the lower 

A synchonous detector is basically a product detector.  With the product 
detector, the only output signal components are the beat note products 
between the carrier and the sidebands, and the carrier and the noise 
components.  Theoretically, there are no beat components between the noise 
and the sideband components, or between various noise components.  The 
result is a better signal to noise ratio because of the absence of all the 
intermodulation of sideband, carrier and noise components that occurs with 
an envelope detector, such as the conventional diode detector.

The problem with attempting to copy AM using a SSB type product detector is 
that in order for the demodulated upper and lower sideband components to 
vectorially add in the proper manner, the re-inserted carrier must not only 
be on the exact  frequency as the original AM carrier; it must be exactly in 
phase or exactly 180 degrees out of phase.   In other words, the reinserted 
carrier of the BFO must be in synchronism with the original AM  carrier. One 
way this can be accomplished is by using PLL technology.

The vectorial addition of the components of both sidebands actually gives 
the demodulated signal an additional 6 dB boost in amplitude. But because 
the DSB signal requires twice the receiver bandpass, twice the background 
noise (3 more dB) is admitted, so the actual boost in s/n ratio is 3 dB 
instead of 6.  So with a product detector, the AM signal should be at least 
as readable in the noise, if not more, than a SSB signal of equivalent 
sideband power.  The majority of the "advantage" that SSB is said to enjoy 
over AM is due to the fact that SSB is demodulated with a more efficient 
detector, the product detector.  With a proper synchonous product detector, 
the SSB advantage is far less.  In addition, the carrier of the AM signal 
can be  reduced or even suppressed, since it is the re-inserted carrier that 
demodulates the sidebands.  With the envelope detector, the original AM 
carrier has to do all the work.  When the original carrier fades, you  get 
selective fading distortion with the envelope detector.

A particular configuration of the synchonous detector, called the Costas 
Loop, can determine the frequency/phase of the original carrier  strictly 
from phase relationships between the two coherent sidebands, without using 
the original carrier as a reference.  Double-sideband suppressed carrier can 
give a 3 dB improvement in s/n ratio over single-sideband suppressed 
carrier, running the same total sideband power.

A PLL synchronous detector that uses the original AM carrier as reference 
can allow for "exalted  carrier" reception which allows DSB AM to function 
at least at least 3 dB better, under noise conditions, as a SSB signal with 
equivalent total sideband power.  This type of detector works with DSB 
reduced carrior AM.  The AM  carrier can be reduced without being totally 
suppressed, as long as the residual, or "pilot" carrier is sufficent for the 
synchronous detector to latch onto. 

More information about the AMRadio mailing list

This page last updated 18 Oct 2017.