Anyone have access to a VHF spectrum analyzer?

What spec an bandwidth are you running please?

The 19 kHz pilot is just that. A steady, pure 19 kHz sinewave. It is doubled to 38 kHz before being used to decode the stereo signals.

--
http://improve-usenet.org/index.html

aioe.org, Goggle Groups, and Web TV users must request to be white
listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:
The crazy, and the insane.
The first sign of insanity is denying that you\'re crazy.

That's true after demodulation.

Looked at in the RF spectrum, though, the situation is going to be a bit different. If you see an RF sideband out 19 kHz from the main carrier, it could be partly from the 19 kHz baseband pilot tone, and partially from the second sideband of a 9.5 kHz component of the audio signal, and/or from the third sideband of a 6.333 kHz component of the audio... etc.

I don't think you can distinguish these cases just by looking at the narrow slice of the RF spectrum around 19 kHz offset from the carrier. It isn't a matter of how narrowly you set your spectrum-analysis filters, since there can be a "legitimate" sideband at 19 kHz offset coming from any number of components in the audio baseband signal.

I think that the only good way to assess the strength of the pilot tone is to do the FM demodulation and then use a narrow-band filter on the demodulated signal.

Strength-of-pilot-tone != strength-of-sideband-at-tone-offset-frequency.

It's certainly possible that some FM stations have a problem, in that the strength of their pilot signal (after demodulation) drops during periods of high audio-signal modulation. This would be a version of the "talk-off" problem. I'd think it would be a poor design of the modulator/transmitter to do audio AGC or clipping *after* addition of the 38 kHz stereo subchannel and 19 kHz pilot, rather than before!

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!

Turns out my pics didn't answer his (unstated) question. And I don't see the stereo subcarrier lines anyhow. Still, it's interesting. We're in plain sight of the Sutro Tower antennas, so the RF field is intense and the FM spectrum is just about full. I think I could run an LED from a small antenna.

John

I would think a prescribed 15KHz low pass filter takes care of that pretty well.

Don't you agree?

M

"Grant" = another incorrigible TOP POSTER

** Wot a pile of putrid, circumlocuting verbal diarrhoea.

This guy must be budding project manager.

** Heavily frequency modulated by an unpredictable audio signal.

Fool.

..... Phil

Interesting. Stereo FM stations broadcast a 19 kHz tone which is used for two purposes: 1) to operate the receiver's stereo indicator, and 2) as a reference for the receiver to double to 38 kHz and use as an LO to demodulate the upper audio bands.

Now it occurs to me that I have never seen a receiver's stereo indicator flicker (get talked off) during loud audio conditions. One would think that the addition and cancellation effect of audio sidebands right at 19 kHz should make the pilot tone flaky, causing flicker of the indicator.

I don't think this can be explained by saying that the stereo indicator has a long time constant to ride through brief pilot interruptions, because I DO see stereo indicator flicker in my car radio during weak signal or multipath conditions.

I think the answer to this may be that the power density of the higher-order audio sidebands at 19 kHz is much less that that of the pilot. The audio is modulated and therefore has low spectral density, whereas the pilot is a CW tone and injected at a relatively strong (10%) level. On the other hand, a loud audio tone at 9500 Hz would have a strong sideband at 19 kHz and probably WOULD cause flicker. However, this would have to be proven using a test FM modulator.

What are your thoughts?

Make a bunch with white LEDs and light your whole building. ;-)

--
http://improve-usenet.org/index.html

aioe.org, Goggle Groups, and Web TV users must request to be white
listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:
The crazy, and the insane.
The first sign of insanity is denying that you\'re crazy.

Nope... because the indicator isn't looking at what you think it's looking at :-)

You're right - it's not a matter of the time constant.

The explanation is simpler than that.

Stereo tuners do *not* look at the 19-kHz-away-from-the-carrier RF sideband in order to detect the presence of stereo.

Instead, they look for a 19 kHz tone in the *demodulated* audio signal (after applying a boatload of gain, limiting, and the actual frequency-to-amplitude detection). The demodulated signal doesn't have the (apparent) ambiguity and cancellation effect that you're concerned about.

Actually, the amplitudes / spectral densities don't matter at all to the system, as long as you aren't generating nonlinear distortion in the audio mixing, or applying so much total modulation that you force the carrier to deviate outside of its permitted range.

It would, *if* the receiver were looking directly at the RF sideband near the original carrier frequency. However, it isn't done that way, for two reasons: [1] It wouldn't work right, for the reasons you indicate, and [2] it would be more difficult and expensive to do it that way.

The effect you indicate doesn't happen, in the actual way of doing things, because the FM demodulation process converts the (complex and overlapping) set of additive sidebands back to the original audio baseband signal plus the pure pilot tone. The audio content and pilot don't overlap or interfere, in a properly-demodulated stereo FM signal, because the audio signal was lowpass-filtered down to around

15 kHz in bandwidth... even though their FM sidebands *did* overlap!

FM modulation and demodulation are somewhat counter-intuitive. A pure single-frequency sinewave intelligence signal, FM-modulated onto a carrier, creates a theoretically-infinite set of sidebands in the RF domain... and the demodulation converts this complex set of phased sidebands back to a single pure tone at the receiver.

If you FM-modulate two (or more) pure tones onto a single carrier, each modulated tone creates a bunch of evenly-space sidebands, which may add or subtract if they happen to fall on the same frequencies... and yet when you demodulate, you end up with the original tones in the original amplitudes, with no loss from the sideband addition or cancellation.

Yeah, it's weird and wonderful.

Nope.

What you say would be true if we were speaking about AM modulation. In AM, each baseband signal creates precisely one pair of sidebands, equally spaced on either side of the carrier by the frequency of the baseband signal. The amplitude of each sideband is directly proportional to the amplitude of the baseband signal.

FM doesn't work that way. With FM, a single pure tone in the baseband creates a theoretically *infinite* set of sidebands on either side of the carrier. The amplitudes of these sidebands are different (at any given level of modulation) and vary according to the values of a series of Bessel functions. In fact, as you increase the amplitude of the baseband signal, the amplitude of the first pair of sidebands rises up to a maximum, and then falls back to zero(!) before reappearing.

If you FM-modulate a carrier with (e.g.) a perfectly pure 1 kHz audio tone, the resulting FM spectrum will consist of the carrier, plus multiple sidebands at offsets of 1 kHz, 2 kHz, 3 kHz, ... for as far as you want to go out looking (although the amplitudes of the further-out sidebands drop down into the noise level and can be neglected).

Low-pass-filtering the signal prior to modulation does *not* eliminate these far-reaching sidebands.

Demodulating the signal accurately (with low distortion) requires that your receiver have a passband which includes all of these sidebands. If you chop off the outer sidebands prior to demodulating (with e.g. a too-narrow IF filter) you'll find that the louder portions of your audio signal become distorted.

So, in the case I was talking about, a 6.333 kHz audio signal (even if perfectly distortion-free, and fed through a 15 kHz low-pass filter) will, when modulated, result in RF sidebands at 6.333 and 12.666 and

19.000 and 25.333 and ... kHz away from the carrier. One of these will fall right on top of the first sideband from the 19 kHz pilot tone.

Weird but true. The books say it works this way, the math says that it works this way, and you can actually see it happen if you've got an RF signal generator with an FM modulator, an audio generator to give you variable-amplitude test tones, and a spectrum analyzer.

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!

Dave, my post above is proof that I can't do two things (talk on the phone and post intelligently) at the same time. I was writing and thinking about two completely different things as you rightly pointed out. Sorry for the waste of time.

I believe the direct 19 kHz detection at RF may work because of the spectral density advantage of the tone vs. the noise-like program audio interfering with it. I can afford the time to wait for a good narrowband phase lock on the tone and average out transient program noise. Program sidebands will be present at 19 kHz but should average out. Also, program audio power (despite pre-emphasis) tends to be greatest is at the lower audio frequencies, so the Bessel components up at 19 kHz will be 20th-order or more, with lower amplitude. So I'm thinking my 19 kHz detector and PLL loop filter can track this out. But, there's technical and schedule risk in trying to do this.

Instead I'll look at using a threshold extension demod for this application. I had been hoping to avoid FM demodulation because of the threshold effect.

Find a Ultrasonic spectrum analyzer an look at the pilot of raw, FM stereo. As a former broadcast engineer, I have never seen a change in that signal, unless the stereo generator was badly damaged. A crystal controlled 38 KHz source is used to create the stereo signals. It is divided by two, and filtered, then fed to the output with the stereo information. It is a constant level. If the received signal is strong enough for hard limiting, the pilot signal will not vary.

BTW, BTSC, or NTSC stereo TV used the horizontal signal for the pilot, and created the stereo signal with 2X horizontal.

The early stereo decoders were very simple. I could dig up some schematics, if I have to.

* * * * * * * Here is the directory of CFR 47 which covers communications:

* * * * * * *

Here are the parts of the FCC documents concerning stereo:

FM Stereo

* * * * * * *

TV Stereo

* * * * * * *

Stereophonic pilot subcarrier use during monophonic programming.

* * * * * * *

multiplex

* * * * * * *

--
http://improve-usenet.org/index.html

aioe.org, Goggle Groups, and Web TV users must request to be white
listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:
The crazy, and the insane.
The first sign of insanity is denying that you\'re crazy.

Grant,

I posted some spectrum analyzer screenshots over on ABSE.

I'm still skeptical that there's an easier way to examine the stereo pilot than to just FM demodulate and then look at 19kHz in the baseband.... especially given that once you've gone to the effort to build a PLL to track an FM carrier, the actual demodulatoin is pretty straightforward -- particularly if you're doing it digitally.

Best wishes,

---Joel

I'm guessing that's alt.binaries.schematics.electronics. My newsgroup server doesn't have that NG.

Found them via a Web feed. The bandwidths are too wide to see the 19 kHz sidebands, though.

lot

-- =A0

Gant,

come on you have to share with us now...what is your application for detecting the presense of a stereo pilot tone on an FM signal that is too noisy to demodulate?

Mark

Grant,

come on you have to share with us now...what is your application for detecting the presense of a stereo pilot tone on an FM signal that is too noisy to demodulate?

Mark

** You are the only person to correctly describe why the OP's daffy notion is totally wrong. .

He also knows that now too and that the game is up.

But watch how he lives on in complete denial - to protect his bloated ego, of course.

Funny to watch as the fool he digs his hole deeper, inch by inch.

He has just gotta be a code scribbler.

..... Phil

lot

-- =A0

Gant, I googled "fm spectra two tone" and this poped up a ways down the list

See Fig 1, note that Fc all the way on the left is the CARRIER.

Mark

If you look at the CFR 47 links I posted you'll see that the 19 kHz signal is under 8% of the total power. Also, the transmission is FM, but the stereo or any subcarrier services aren't.

--

aioe.org, Goggle Groups, and Web TV users must request to be white listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in your account:

There are two kinds of people on this earth: The crazy, and the insane. The first sign of insanity is denying that you're crazy.