"David" wrote in message
...
The only thing that keeps us from hearing the digital sidebands is
their symmetry (the digital sidebands are out of phase with each other
and cancel out in the receiver).

I'm still trying to figure out how this is supposed to work.. I understand
the concept, FM stereo does something similar by using phasing to reduce the
bandwidth needed for the L-R signal. The peaks of the L-R fit nicely into
the valleys of the L+R signal and vice versa.

But this doesn't seem like it would work as advertised on a store-stock AM
radio, because the ordinary envelope detector does not detect both
sidebands, only one of them. So how does it cancel?

Brenda Ann wrote:


"David" wrote in message
...
The only thing that keeps us from hearing the digital sidebands is
their symmetry (the digital sidebands are out of phase with each other
and cancel out in the receiver).

I'm still trying to figure out how this is supposed to work.. I understand
the concept, FM stereo does something similar by using phasing to reduce
the bandwidth needed for the L-R signal. The peaks of the L-R fit nicely
into the valleys of the L+R signal and vice versa.

No, this is not how FM works. I suggest you do some google searching. Here
is a start.
http://www.radio-electronics.com/inf...modulation.php
http://www.st-andrews.ac.uk/~www_pa/...t13/page1.html
http://www.fas.org/man/dod-101/navy/docs/es310/FM.htm


But this doesn't seem like it would work as advertised on a store-stock AM
radio, because the ordinary envelope detector does not detect both
sidebands, only one of them. So how does it cancel?


An envelope detector does just what its name implies. It works from the
peaks in the time domain signal. The signal applied to the detector is
everything that comes through the filter; the carrier and both sidebands if
the signal is centered in the filters passband. The only way you can
exclude a sideband is to tune off from the center frequency.

A little bit of google searching can find you these.
http://en.wikipedia.org/wiki/Envelope_detector
http://www.radio-electronics.com/inf...e_am_demod.php
http://www.st-andrews.ac.uk/~jcgl/Sc...rt9/page2.html
http://www.mit.bme.hu/research/chaos...1_04am_dem.pdf

"craigm" wrote in message
...
Brenda Ann wrote:


"David" wrote in message
...
The only thing that keeps us from hearing the digital sidebands is
their symmetry (the digital sidebands are out of phase with each other
and cancel out in the receiver).

I'm still trying to figure out how this is supposed to work.. I
understand
the concept, FM stereo does something similar by using phasing to reduce
the bandwidth needed for the L-R signal. The peaks of the L-R fit nicely
into the valleys of the L+R signal and vice versa.

No, this is not how FM works. I suggest you do some google searching. Here
is a start.

You need to check out the math and the practise of FM Stereo MPX. It is as I
have stated previously.

But this doesn't seem like it would work as advertised on a store-stock
AM
radio, because the ordinary envelope detector does not detect both
sidebands, only one of them. So how does it cancel?


An envelope detector does just what its name implies. It works from the
peaks in the time domain signal. The signal applied to the detector is
everything that comes through the filter; the carrier and both sidebands
if
the signal is centered in the filters passband. The only way you can
exclude a sideband is to tune off from the center frequency.

The detector in your basic AM radio is much the same as it has been for
nearly 100 years now. It rectifies one half of the envelope and filters out
the remaining RF to leave an audio waveform. It does not detect both halves
(both sidebands) of the waveform. This is why such things as selectable
sideband make high end radios better able to pick out a signal. The signal
with both sidebands may be applied to the detector, but it's not what comes
out. I stand by my question. If only one sideband is actually detected,
there can be no phase cancellation.

Brenda Ann wrote:


"craigm" wrote in message
...
Brenda Ann wrote:


"David" wrote in message
...
The only thing that keeps us from hearing the digital sidebands is
their symmetry (the digital sidebands are out of phase with each other
and cancel out in the receiver).

I'm still trying to figure out how this is supposed to work.. I
understand
the concept, FM stereo does something similar by using phasing to reduce
the bandwidth needed for the L-R signal. The peaks of the L-R fit nicely
into the valleys of the L+R signal and vice versa.

No, this is not how FM works. I suggest you do some google searching.
Here is a start.

You need to check out the math and the practise of FM Stereo MPX. It is as
I have stated previously.

But this doesn't seem like it would work as advertised on a store-stock
AM
radio, because the ordinary envelope detector does not detect both
sidebands, only one of them. So how does it cancel?


An envelope detector does just what its name implies. It works from the
peaks in the time domain signal. The signal applied to the detector is
everything that comes through the filter; the carrier and both sidebands
if
the signal is centered in the filters passband. The only way you can
exclude a sideband is to tune off from the center frequency.

The detector in your basic AM radio is much the same as it has been for
nearly 100 years now. It rectifies one half of the envelope and filters
out
the remaining RF to leave an audio waveform. It does not detect both
halves (both sidebands) of the waveform. This is why such things as
selectable
sideband make high end radios better able to pick out a signal. The
signal with both sidebands may be applied to the detector, but it's not
what comes out. I stand by my question. If only one sideband is actually
detected, there can be no phase cancellation.


You clearly don't understand how an AM detector works.

"craigm" wrote in message
...
The detector in your basic AM radio is much the same as it has been for
nearly 100 years now. It rectifies one half of the envelope and filters
out
the remaining RF to leave an audio waveform. It does not detect both
halves (both sidebands) of the waveform. This is why such things as
selectable
sideband make high end radios better able to pick out a signal. The
signal with both sidebands may be applied to the detector, but it's not
what comes out. I stand by my question. If only one sideband is actually
detected, there can be no phase cancellation.


You clearly don't understand how an AM detector works.


But I do. The basic AM detector works exactly as any half-wave rectifier. If
you used a full-wave rectification, then you could detect both sides of the
RF waveform, but logic would seem to dictate that this would cancel the
audio waveform, since the two sidebands are mirror images of one another.

Brenda Ann wrote:


"craigm" wrote in message
...
The detector in your basic AM radio is much the same as it has been for
nearly 100 years now. It rectifies one half of the envelope and filters
out
the remaining RF to leave an audio waveform. It does not detect both
halves (both sidebands) of the waveform. This is why such things as
selectable
sideband make high end radios better able to pick out a signal. The
signal with both sidebands may be applied to the detector, but it's not
what comes out. I stand by my question. If only one sideband is actually
detected, there can be no phase cancellation.


You clearly don't understand how an AM detector works.


But I do. The basic AM detector works exactly as any half-wave rectifier.

That is correct.

If you used a full-wave rectification, then you could detect both sides of
the RF waveform,

So far, correct.

but logic would seem to dictate that this would cancel
the audio waveform,

No cancellation.

since the two sidebands are mirror images of one
another.

In the frequency domain, yes, that is true. But an envelope detector works
in the time domain. The two are not interchangeable.

You seem to believe that the positive peaks of the AM waveform represent one
sideband and the negative peaks, the other. That is grossly incorrect.

If all you need to add is a diode to detect the opposite sideband, it would
be cheap and easy for all radios to be sideband selectable with the flip of
a switch. But you don't see that, ever wonder why?

If the diode based envelope detector only detected one sideband, then half
the radios in the world would have trouble receiving CHU as it does not
broadcast one sideband. (Adding a balun and getting the wires reversed
would invert the signal and you could lose reception.) This follows from
what you have suggested, yet it doesn't happen.

Look at the links I provided.

If a full wave rectifier is used, the AM envelope is updated twice as often.
It can be viewed as flipping the bottom half of the waveform up and using a
single diode. (Look at the waveforms for a power supply with half wave and
full wave rectification and this may be clearer.)

To receive a single sideband, you would need to tune the radio so that the
carrier is at one edge of the passband of the filter. Then one sideband
will be reduced somewhat, depending upon the shape of the filter, and the
other sideband will be in the middle of the passband. This generally will
not give good rejection of the unwanted sideband.

The other approach is to use the phasing detector as would be found in some
(sideband selectable) synchronous detectors or some SSB detectors. These
detectors process both sidebands and use cancellation to remove the
unwanted sideband.

On Sep 15, 10:05 pm, "Brenda Ann" wrote:
"craigm" wrote in message

...

The detector in your basic AM radio is much the same as it has been for
nearly 100 years now. It rectifies one half of the envelope and filters
out
the remaining RF to leave an audio waveform. It does not detect both
halves (both sidebands) of the waveform. This is why such things as
selectable
sideband make high end radios better able to pick out a signal. The
signal with both sidebands may be applied to the detector, but it's not
what comes out. I stand by my question. If only one sideband is actually
detected, there can be no phase cancellation.

You clearly don't understand how an AM detector works.

But I do. The basic AM detector works exactly as any half-wave rectifier. If
you used a full-wave rectification, then you could detect both sides of the
RF waveform, but logic would seem to dictate that this would cancel the
audio waveform, since the two sidebands are mirror images of one another.

You are mistaken in thinking that the positive and negative sides of
the RF envelope correspond to the upper and lower sidebands. They do
not. Examine a SSB envelope - it has positive and negative components
that must balance each other, else there would be a DC component. How
could DC be delivered through a vacuum? It is correct that applying
modulation to one side of an Independent Sideband transmitter and anti-
phase modulation to the other side will result in cancellation in the
envelope detector, whether it is half-wave or full-wave, provided both
sidebands are delivered equally. If one sideband is cut off (by
filtering in the receiver), then there would be no cancellation. For
normal DSB-AM, the two sidebands are modulated in phase and the
envelope will be detected by both half-wave and full-wave.detectors,
whether one sideband or both are passed to the detector with carrier.

Regards,
Tom

"Frank Dresser" wrote in message
...

"craigm" wrote in message
...

[snip]

You seem to believe that the positive peaks of the AM waveform represent
one
sideband and the negative peaks, the other. That is grossly incorrect.

If all you need to add is a diode to detect the opposite sideband, it
would
be cheap and easy for all radios to be sideband selectable with the flip
of
a switch. But you don't see that, ever wonder why?


[snip]

I read about such an AM stereo system in an old radio magazine. The
positive peaks of the envelope carried one channel, the negative peaks
carried the other channel. The idea was patented, but I doubt it was
developed to any extent. The normal "flywheel effect" of the L-C
filtering
would symmetrize the modulation envelope and destroy the stereo
seperation.
The higher the audio frequency, the the poorer the stereo seperation would
be. This system was completely compatable with normal AM radios but
adequate stereo seperation would have required such low Q circuits that
these AM stereo radios would receive way too much noise and interference.

And, as we've seen, it's far better to try and sell magic radios when
everybody else has to put up with the noise and interference.

Frank Dresser

That would be the original Kahn system. I forget the exact name of it. XETRA
used it for many years, and if you had two radios you could tune off channel
high on one and low on the other and get stereo separation. I listened to it
occasionally. It wasn't great, but it was an interesting novelty at the
time.

"Brenda Ann" wrote in message
...

That would be the original Kahn system. I forget the exact name of it.
XETRA
used it for many years, and if you had two radios you could tune off
channel
high on one and low on the other and get stereo separation. I listened to
it
occasionally. It wasn't great, but it was an interesting novelty at the
time.




Could you get both left and right audio channels when using two radios which
had their detector diodes installed with the same polarity?

Frank Dresser