How does PCM-Demodulation work?

On Apr 28, 2:30 am, "Green Xenon [Radium]" <gluceg...@excite.com>
wrote:
> glen herrmannsfeldt wrote:
> > Randy Yates wrote:
> > (snip)
>
> >> I'm guessing so. In almost 30 years, I've never heard of a
> >> PCM-demodulator, but I am getting old and senile. Where did you come
> >> across this term?
>
> > Well, the M is for modulation, but I agree using the term
> > "demodulator" in that case is unusual.
>
> > In the past, other modulation methods such as PWM (pulse width
> > modulation) were used.  In that case, a real demodulator makes
> > some sense.  Compare to demodulating an AM-SC (Amplitude
> > modulation - suppressed carrier) signal like the stereo
> > subcarrier.
>
> > For PWM, even as a 'digital' signal, the demodulation is
> > somewhat analog.  For PWM, you might have a PLL to lock onto
> > the bit stream (such as that coming off a CD), but the actual
> > demodulation is the D/A converter.
>
> > -- glen
>
> A PCM-demodulator does to a PCM-signal what an AM-demodulator does to an
> AM-signal which is also what an FM-demodulator does to an FM-signal. I
> know how AM/FM demodulation work. I am just curious as to how
> PCM-demodulation works. From that start, I've also believed that
> PCM-demodulation has nothing to do the DAC conversion at all.

Why do you say it has *nothing* to do with the DAC conversion?  The
"modulation" of PCM is the ADC step.  The inverse of ADC is DAC so why
wouldn't that be the "demodulation" step?

I've always thought of PCM as a very odd term.  There is no modulation
involved.  Calling it "coding" might be ok, but why "pulse"?  There
are not really pulses involved, just digital data.  But if you
consider it in the context of the times when it was being used, the
other choices were all analog modulation techniques; AM, FM, PWM,
PAM,...  In that context, it made sense to call digital data a "pulse
modulation" of some sort since it is digital which is always pulses
(compared to analog signals).

PCM is just a name for digital representation of an analog signal.
The name is not descriptive or accurate and it does not fit "proper"
nomenclature.  But it is with us to stay.  Don't try to read too much
into it.

Rick

rickman wrote:
...
> I've always thought of PCM as a very odd term.  There is no modulation
> involved.  Calling it "coding" might be ok, but why "pulse"?  There
> are not really pulses involved, just digital data.  But if you
> consider it in the context of the times when it was being used, the
> other choices were all analog modulation techniques; AM, FM, PWM,
> PAM,...  In that context, it made sense to call digital data a "pulse
> modulation" of some sort since it is digital which is always pulses
> (compared to analog signals).
> 


The BBC used the term PCM for their early digital audio transmissions. 
There were in effect two stages -  the initial ADC stage, and a further 
stage to "encode" it for broadcast or relay, e.g. as part of the overall 
video feed. From what little I remember, this was not so much for direct 
broadcast to users (so not "digital radio" as such), but to transmit 
from location to studio etc (e.g. over microwave link). The data was not 
"pure" ADC output, but carried additional parity bits and stuff (to use 
the full technical term). Googling "PCM BBC" should throw up lots of 
relevant material.

Richard Dobson

Green Xenon [Radium] wrote:

> ... I've also believed that 
> PCM-demodulation has nothing to do the DAC conversion at all.

You may remember that you've been wrong before.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

rickman wrote:

(snip regarding PCM, and modulation in general)

> Why do you say it has *nothing* to do with the DAC conversion?  The
> "modulation" of PCM is the ADC step.  The inverse of ADC is DAC so why
> wouldn't that be the "demodulation" step?

Well, yes, but it is also in the storage method for the
ADC output.

> I've always thought of PCM as a very odd term.  There is no modulation
> involved.  Calling it "coding" might be ok, but why "pulse"?  There
> are not really pulses involved, just digital data.  But if you
> consider it in the context of the times when it was being used, the
> other choices were all analog modulation techniques; AM, FM, PWM,
> PAM,...  In that context, it made sense to call digital data a "pulse
> modulation" of some sort since it is digital which is always pulses
> (compared to analog signals).

It is just digital data, but you have to do something with
it to store it or send it through a communications link.

In pretty much all systems the digital data goes through analog
electronics and is stored as a modulated signal.  (I would
possibly give an exception to an NRZ signal through a DC coupled
system.)

> PCM is just a name for digital representation of an analog signal.
> The name is not descriptive or accurate and it does not fit "proper"
> nomenclature.  But it is with us to stay.  Don't try to read too much
> into it.

Consider what it actually looks like on a CD, or the signal coming
off the CD.  It it recorded as on/off pulses of the laser.

Before CD there were other optical recording systems
being considered, such as PWM.  PWM is 'digital' in that the
signal is either on or off, and so not dependent on the linearity
of the recording/playback system.  PWM could also be time
multiplexed on an analog medium, such as optical fiber or
coaxial cable.   Pulse modulated signals can go through AC
coupled systems such as transformers or magnetic tape.

But yes, the term PCM is often used in the context where
no modulation is needed.  (Though stored on a magnetic disk
on a computer it is also pulse modulated in some form.)

-- glen