FM stereo
RichD
It has a FM stereo detector, with internal
threshold. When the signal gets weak or
noisy, it switches to mono, and the noise
drops significantly.
How does FM stereo decoding work, and
why is it so much noisier than mono?
--
Rich
Phil Allison
Joel Koltner
news:1e37a98c-ad31-48d4...@l35g2000vba.googlegroups.com...
> How does FM stereo decoding work, and
> why is it so much noisier than mono?
The short answer is that the vast bulk of the power being transmitted goes to
L+R using regular old FM as you see any textbook describe. When stereo was
added, they just stuck a subcarrier at 38kHz in the baseband (which is ignored
by monoaural receivers that cut off at 15kHz of thereabouts) and encoded L-R
nearby... using a small fraction of the total signal power.
Since you need L-R to recover the left and right channels separately (L+R
added to L-R gets you L; subtract the two and you get R), as the station's
signal gets weaker the SNR of L-R drops into the noise more quickly than that
of L+R and hence stereo broadcasts become noisy even when monaural (L+R) is
still doing OK.
Wikipedia or Google can find quantitative results for things like "vast bulk
of the power" if you're so inclined. :-)
---Joel
Joerg
It's here:
http://en.wikipedia.org/wiki/FM_broadcasting
The pilot tone is being decoded and when its level is to the receiver's
liking it switches to stereo.
--
Regards, Joerg
http://www.analogconsultants.com/
"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Phil Allison
"Joerg"
> The pilot tone is being decoded
** The 19kHz pilot tone contains no information - it's purely a reference
of frequency and phase for the stereo decoder.
> and when its level is to the receiver's liking it switches to stereo.
** The level of the pilot tone is fixed at 10% max modulation of the
carrier.
Automatic stereo/mono switching is carried out on the basis of the signal
strength in the IF stage.
.... Phil
Joerg
The last ones I've dealt with used the SNR of the pilot tone to do
exactly that :-)
Phil Allison
"Joerg"
> Phil Allison wrote:
>> "Joerg"
>>
>>> The pilot tone is being decoded
>>
>> ** The 19kHz pilot tone contains no information - it's purely a
>> reference of frequency and phase for the stereo decoder.
>>
>>
>>> and when its level is to the receiver's liking it switches to stereo.
>>
>>
>> ** The level of the pilot tone is fixed at 10% max modulation of the
>> carrier.
>>
>> Automatic stereo/mono switching is carried out on the basis of the signal
>> strength in the IF stage.
>>
>
> The last ones I've dealt with used the SNR of the pilot tone to do exactly
> that :-)
** Utter BOLLOCKS.
Piss off - WOG MORON.
..... Phil
Joerg
Oh yeah?
http://www.st.com/stonline/books/pdf/docs/1556.pdf
Quote "The pilot detector output is designed as an open collector
output, therefore an external pullup resistor is
needed. To force the decoder to "MONO" Pin 19 has to be clamped to a
voltage below 0.8V."
Phil Allison
"Joerg" = fuckwit WOG from HELL
>>>>> The pilot tone is being decoded
>>>
>>>> ** The 19kHz pilot tone contains no information - it's purely a
>>>> reference of frequency and phase for the stereo decoder.
>>>>
>>>>
>>>>> and when its level is to the receiver's liking it switches to stereo.
>>>>
>>>> ** The level of the pilot tone is fixed at 10% max modulation of the
>>>> carrier.
>>>>
>>>> Automatic stereo/mono switching is carried out on the basis of the
>>>> signal strength in the IF stage.
>>>>
>>> The last ones I've dealt with used the SNR of the pilot tone to do
>>> exactly that :-)
>>
>>
>> ** Utter BOLLOCKS.
>>
>
> Oh yeah?
>
> http://www.st.com/stonline/books/pdf/docs/1556.pdf
>
> Quote "The pilot detector output is designed as an open collector output,
> therefore an external pullup resistor is
> needed. To force the decoder to "MONO" Pin 19 has to be clamped to a
> voltage below 0.8V."
** Giant HUH ???
This anencephalic must be posting from a mental asylum.
FUCK off - YOU WOG MORON.
.... Phil
mi...@sushi.com
I see Phil is showing off his intelligence again. ;-)
Jan Panteltje
<r_dela...@yahoo.com> wrote in
<1e37a98c-ad31-48d4...@l35g2000vba.googlegroups.com>:
FM encoding uses a 38kHz carrier on which the L-R and R-R signals are QAM modulated.
To sync the carrier, a 19 kHz (f/2) pilot tone (above audible) is transmitted with the audio.
The audio is normally transmitted as L+R.
After decoding, the signals can be combined by simple addition into L and R.
The noise decreases when you switch to mono, because the
whole decoder is switched off, and that part of the spectrum that contains noise,
the sidebands around 38 kHz, is no longer fed into the signal matrix.
All this IIRC.
Jan Panteltje
<r_dela...@yahoo.com> wrote in
<1e37a98c-ad31-48d4...@l35g2000vba.googlegroups.com>:
>I have a digital armband radio, for jogging.
FM encoding uses a 38kHz carrier on which the L-R and R-R signals are QAM modulated.
To sync the carrier, a 19 kHz (f/2) pilot tone (above audible) is transmitted with the audio.
The audio is normally transmitted as L+R.
After decoding, the signals can be combined by simple addition into L and R.
The noise decreases when you switch to mono, because the
whole decoder is switched off, and that part of the spectrum that contains noise,
the sidebands around 38 kHz, is no longer fed into the signal matrix.
All this IIRC.
PS:.
I did some experiments with L+R and L-R encoding in mp3 and AC3, mainly originating
from a request for a very narrow bandwidth audio link that would drop
to mono if even less bandwidth was available:
http://panteltje.com/panteltje/dvd/mono-stereo.txt
So you can have a mono mp3, that is L+R, and then send L-R over a second channel,
even with a different, or LESS, bandwidth,
and at the other end combine the signals, and have good stereo again.
And if the second L-R channel fails, it will just drop back to mono.
Works very well, even with [in] highly compressed formats.
The 'subtract_wave' program I wrote to do this is here:
http://panteltje.com/panteltje/dvd/substract_wave-0.3.tgz
All Linux GPL code.
christofire
"Jan Panteltje" <pNaonSt...@yahoo.com> wrote in message
news:hegkep$2pg$1...@news.albasani.net...
YRNC it would seem! There's no QAM in the commonplace Zenith-GE pilot-tone
system, just DSB-SC AM of a 38 kHz tone by the L-R difference signal. The
L+R sum signal is provided by the monaural, baseband FM signal. L and R are
separated by simultaneous equations and the 19 kHz pilot tone is obtained by
dividing the frequency of 38 kHz tone by two.
Chris
Jan Panteltje
<chris...@btinternet.com> wrote in
<JLWdnfXRicb535HW...@bt.com>:
You are right, it is double sideband surpressed carrier.
>The L+R sum signal is provided by the monaural, baseband FM signal. L and R are
>separated by simultaneous equations and the 19 kHz pilot tone is obtained by
>dividing the frequency of 38 kHz tone by two.
Yes, but in the decoder the 38 kHz is obtained by multiplying the 19 kHz by 2 I think,
and that 38 kHz is then used to demodulate that DSBSC signal.
This right?
christofire
... hence DSB-SC!
>
>
>>The L+R sum signal is provided by the monaural, baseband FM signal. L and
>>R are
>>separated by simultaneous equations and the 19 kHz pilot tone is obtained
>>by
>>dividing the frequency of 38 kHz tone by two.
>
> Yes, but in the decoder the 38 kHz is obtained by multiplying the 19 kHz
> by 2 I think,
> and that 38 kHz is then used to demodulate that DSBSC signal.
> This right?
Yes, the obvious way.
Chris
JosephKK
wrote:
>Phil Allison wrote:
>> "Joerg"
>>
>>> The pilot tone is being decoded
>>
>> ** The 19kHz pilot tone contains no information - it's purely a reference
>> of frequency and phase for the stereo decoder.
>>
>>
>>> and when its level is to the receiver's liking it switches to stereo.
>>
>>
>> ** The level of the pilot tone is fixed at 10% max modulation of the
>> carrier.
>>
>> Automatic stereo/mono switching is carried out on the basis of the signal
>> strength in the IF stage.
>>
>
>The last ones I've dealt with used the SNR of the pilot tone to do
>exactly that :-)
Alas, Phil A. is full of baloney on this one. The stereo decoder
makes that decision, and it cares about S/N of the pilot tone itself.
I have built many with my own hands. Even designed one or two.
JosephKK
wrote:
>Phil Allison wrote:
>> "Joerg"
>>> Phil Allison wrote:
>>>> "Joerg"
>>>>
>>>>> The pilot tone is being decoded
>>>> ** The 19kHz pilot tone contains no information - it's purely a
>>>> reference of frequency and phase for the stereo decoder.
>>>>
>>>>
>>>>> and when its level is to the receiver's liking it switches to stereo.
>>>>
>>>> ** The level of the pilot tone is fixed at 10% max modulation of the
>>>> carrier.
>>>>
>>>> Automatic stereo/mono switching is carried out on the basis of the signal
>>>> strength in the IF stage.
>>>>
>>> The last ones I've dealt with used the SNR of the pilot tone to do exactly
>>> that :-)
>>
>>
>> ** Utter BOLLOCKS.
>>
>
>Oh yeah?
>
>http://www.st.com/stonline/books/pdf/docs/1556.pdf
>
>Quote "The pilot detector output is designed as an open collector
>output, therefore an external pullup resistor is
>needed. To force the decoder to "MONO" Pin 19 has to be clamped to a
>voltage below 0.8V."
Oh gosh. You are talking about IC's that do it all for you. I have
done one with tunable "cans" (resonant transformers similar to IF
transformers), and one with a 76 kHz PLL. The PLL performed much
better. About 40 years ago though.
JosephKK
<pNaonSt...@yahoo.com> wrote:
>On a sunny day (Mon, 23 Nov 2009 16:07:30 -0800 (PST)) it happened RichD
><r_dela...@yahoo.com> wrote in
><1e37a98c-ad31-48d4...@l35g2000vba.googlegroups.com>:
>
>>I have a digital armband radio, for jogging.
>>
>>It has a FM stereo detector, with internal
>>threshold. When the signal gets weak or
>>noisy, it switches to mono, and the noise
>>drops significantly.
>>
>>How does FM stereo decoding work, and
>>why is it so much noisier than mono?
>>
>>--
>>Rich
>
>FM encoding uses a 38kHz carrier on which the L-R and R-R signals are QAM modulated.
>To sync the carrier, a 19 kHz (f/2) pilot tone (above audible) is transmitted with the audio.
>The audio is normally transmitted as L+R.
>
Not QAM (at least not originally) just a double balanced modulator
thus no 38 kHz carrier.
>After decoding, the signals can be combined by simple addition into L and R.
>The noise decreases when you switch to mono, because the
>whole decoder is switched off, and that part of the spectrum that contains noise,
>the sidebands around 38 kHz, is no longer fed into the signal matrix.
>All this IIRC.
It was actually set up so that the math worked correctly for TDM
decoding at 38 kHz or FDM subcarrier and matrix decoding. Some
performance tradeoffs were observed later. It turned out that decoder
solutions that used a bit of both techniques had the best performance.
All known 40 years ago.
JosephKK
One obvious way. It however required carefully tuned "cans". The 76
kHz PLL method ended up far simpler and more reliable. It is what all
the ICs implement.
Phil Allison
"JosephKK is CRIMINAL LYING NUT CASE "
Joerg the Wog Moron is his favourite cock sucker
>
>>> The pilot tone is being decoded
>>
>> ** The 19kHz pilot tone contains no information - it's purely a
>> reference
>> of frequency and phase for the stereo decoder.
>>
>>
>>> and when its level is to the receiver's liking it switches to stereo.
>>
>>
>> ** The level of the pilot tone is fixed at 10% max modulation of the
>> carrier.
>>
>> Automatic stereo/mono switching is carried out on the basis of the signal
>> strength in the IF stage.
>>
>
>The last ones I've dealt with used the SNR of the pilot tone to do
>exactly that :-)
>Alas, Phil A. is full of baloney on this one. The stereo decoder
>makes that decision, and it cares about S/N of the pilot tone itself.
** 100% BOLLOCKS.
Decoder ICs have no way of knowing the s/n of the pilot and it is an
undefined thing in any case.
Such ICs operate on the incoming LEVEL of the tone, with some deliberate
hysteresis in the switching.
Being a fixed percentage of the carrier deviation, the pilot tone LEVEL is
fixed !!!
> I have built many with my own hands. Even designed one or two.
** So have I as well - you SMUG CUNTHEAD !!!
Shame YOU are such a fucking STOOPID ASS
you had no clue about what you were dealing with.
FUCK OFF
- YOU RIDICULOUS DAMN TROLL !!!
.... Phil
Paul Keinanen
wrote:
>Decoder ICs have no way of knowing the s/n of the pilot and it is an
>undefined thing in any case.
Why not ?
The mono (L+R) signal extends to 15 kHz, while the lower sideband of
the difference (L-R) starts at 23 kHz extending to 38 kHz, while the
upper sideband goes from 38 to 53 kHz and there may be a data carrier
at 57 kHz and some auxiliary audio (SCA) in some countries.
The noise density from the FM demodulator increases at least with
second power of audio frequency until the IF filter passband again
drops the noise density.
Thus, it sufficient to detect the noise power in some "quiet"
frequency band. For example the 15..23 kHz band is supposed to be
quiet with only the 19 kHz pilot in the middle, so it should be
possible to measure the pilot power and compare it to the total noise
power of the frequency band. Alternatively notch out the 19 kHz and
measure the noise power in that band.
>
>Such ICs operate on the incoming LEVEL of the tone, with some deliberate
>hysteresis in the switching.
>
>Being a fixed percentage of the carrier deviation, the pilot tone LEVEL is
>fixed !!!
As long as the IF amplifier has sufficiently gain to drive the FM
demodulator into saturation at low signal levels, the FM demodulator
pilot tone output level is indeed constant.
However, when the IF amplifier is no longer limiting, the pilot starts
to drop. Even with plenty of IF gain, the FM demodulator is finally
going to be saturated by the noise and hence also the pilot is finally
lost.
However, the dropping of the pilot tone level occurs at such low audio
levels that it usable only for completely muting even the mono
reception and the signal has been useless for stereo reception a long
time ago.
Phil Allison
"Paul Keinanen is a 100% CUNT "
"Phil Allison"
>>Decoder ICs have no way of knowing the s/n of the pilot and it is an
>>undefined thing in any case.
>
> Why not ?
** No circuitry exists inside to do so.
YOU FUCKING TENTH WIT !!
>>Such ICs operate on the incoming LEVEL of the tone, with some deliberate
>>hysteresis in the switching.
>>
>>Being a fixed percentage of the carrier deviation, the pilot tone LEVEL is
>>fixed !!!
>
> As long as the IF amplifier has sufficiently gain to drive the FM
> demodulator into saturation at low signal levels, the FM demodulator
> pilot tone output level is indeed constant.
** Duhhhhhhhhhhhhhh ???
> However, when the IF amplifier is no longer limiting, the pilot starts
> to drop. Even with plenty of IF gain, the FM demodulator is finally
> going to be saturated by the noise and hence also the pilot is finally
> lost.
** Duhhhhhhhhhhhhhhhhhhh ???
> However, the dropping of the pilot tone level occurs at such low audio
> levels
** Wot fucking asinine drivel !!!!!!!!
The audio level not even faintly involved.
> that it usable only for completely muting even the mono
> reception and the signal has been useless for stereo reception a long
> time ago.
** Duuuuuuhhhhhhhhhhhhhhhhhhhhh??
FUCK OFF !!!!!!!!!!!
YOU KNOW NOTHING DAMN TROLL
.... Phil