Anyone have access to a VHF spectrum analyzer?

ff

ox

If you contact the station they will probably tell you what you want.

Leon

The sidebands are constantly changing in number and amplitude with the modulation; but the peak deviation for VHF FM stations is 75 KHz.

"Grant"

** As already pointed out - sideband levels depend on the instantaneous program levels and frequency content.

The largest deviation from centre frequency allowed for commercial FM radio is +/- 75 kHz, while the highest modulating frequency is about 53kHz for stereo FM.

Generally, the whole signal spectrum will be included within +/- 100 kHz of the carrier frequency, with only very small residual levels outside that range.

Frequency dependant program limiting is included in the audio processing chain to make sure this remains the case.

Now, how about you tell us what you ACTUALLY wanted to know ???

..... Phil

See pics in a.b.s.e.

John

What I'd ACTUALLY like to know, is approximately how constant in amplitude the 19 kHz stereo pilot sidebands are when program audio is present. This would mean looking at +/- 19 kHz away from the RF carrier center frequency with a fairly narrow spec an bandwidth to avoid sidebands due to program audio.

AIUI, you can't readily see this on a regular old spectrum analyzer: The 19kHz pilot is in the *demodulated* audio channel, so (as with any pilot tone in the audio channel) it gets spindled, mutilated, and folded into the overall RF spectrum, and unless it's the only thing present (i.e., the audio has been muted), I'm not aware of any easy way to go from looking at an RF spectrum to finding the pilot's level (other than actually performing FM demodulation, of course).

(With only the stereo pilot present, on the RF spectrum you'd see a series of spikes at multiples of 19kHz around the carrier, and you'd then have to consult a Bessel function table or a good calculator to back out what modulation index is and then the pilot level...)

But perhaps there's some old-timer's trick that I'm unaware of here...?

---Joel

"Grant"

** Top post removed to where it should be.

** The 19kHz pilot tone modulates the carrier frequency by no more than 10% of the maximum deviation, ie by around +/- 7 kHz.

The modulating signal is filtered to provide a clear band around 19 kHz ( by

+/- 4 kHz ) so the tone will not be affected by the stereo modulation signal. ** Not true.

You are confusing AM and FM theory.

See:

..... Phil

Exactly right. There are higher-order Bessel components due to program audio at center freq. +/- 19 kHz. Also possible sideband asymmetry, etc. But the 19 kHz unmodulated pilot at 10% injection is a pure tone and will have the highest signal-to-noise density in that part of the spectrum. All I'd like to accomplish for now is to see if it's recognizable (while program audio is present) in a narrow RF bandwidth on a spec analyzer.

Not confused. I'm looking into alternative methods of direct detection at RF of the pilot.

"Grant"

** You are a braver man than I.

FM broadcast signals are hardy ever of steady amplitude at the receiving antenna - unless your receiver is very close to the transmitter. Multi-path signals, moving objects in the environment and weather all affect the carrier signal amplitude.

FM receivers overcome this problem by using HEAVY amplitude limiting in the IF stages, then the FM detector can work unimpeded by amplitude variations.

For your idea to work, you need to amplitude limit the *carrier frequency* and THEN pick out a tiny band, maybe 10 Hz wide just above that carrier to get your steady pilot tone component free of audio modulation components.

Even if you achieve this - what have you got ?

A frequency that is the sum of two crystal time bases, one in the transmitter and another in the FM modulator ( or exciter). Neither of which is anything special in terms of accuracy.

Lotsa luck.

.... Phil

. .

Then why the hell didn't you ask that question to begin with instead of running us around the flagpole with the whole spectrum? And what difference does it make if we take that signal off the air rather than taking the data from the reference specification.

Idiot.

Jim

-- "It is the mark of an educated mind to be able to entertain a thought without accepting it." --Aristotle

on at

ect

n the

s.

*

r to

which

I have a spectrum analyzer that covers the FM band, a stereo generator, and RF source I could modulate. The only thing I don't have is the will do go on a fool's errand. ;-)

I bought the Radiometer 501 generator on a lark. Actually 3 for $60 untested, but only two worked.

tion at

Grant,

I think we talked about this before. I will give you a simple example that demonstrats the problem you need to solve. If the FM station is sending only the 19 kHz pilot and nothing else then you have the 2 sidebands serarated from the carrier by 19 kHz as you estimated. At this point, you could use a narrow band receiver and detect or do whatever you want with those two sidebands. OK so far so good.

But now consider the FM station is also xmitting a bass drum hit say

100 Hz. The main carrier AND THE 19 kHz SIDEBANDS are ALL now going to sweep back and forth at 100 Hz by up to +/- 75 kHz. So if you want to use a narrow band device to detect the 19 kHz sidebands, you need to somehow deal with (track them) the fact that they are sweeping around with the carrier.

I'm not saying it can't be done, but that is the issue you have to solve.

As I told you last time, you want look up the literature about "threshold extension" demodulators.

What is the application that you need to detect the 19kHz pilot tone if the main carrier is so far below threshold?

Mark

OK, I'm kinda leery that there's much "useful" information just looking at what's going on at +/-19kHz in the RF spectrum on an instantaneous basis, but I understand what you're looking for now. I'll take a look at a spectrum analyzer this afternoon and get a screenshot if it shows anything useful...

Thank you Joel. The 19 kHz pilot sidebands will obviously look different depending on what modulation is present in the audio channel. But I'm just hoping to get an idea of how the sidebands appear under "average" audio conditions. Appreciate your taking time to look at this.

Mark, I agree. FM is a non-linear process and in the general case you can't simply add the spectra of individual baseband components at the FM output. In some cases that CAN be done, although those cases involve certain (sinusoidal) modulating signal relationships that don't apply to my particular case.

All I'm interested in for now is seeing how the 19 kHz pilot sidebands appear in the presence of "typical" program audio modulation.

Yes, I'm familiar with TEDs. In fact, I was designing and manufacturing them on the TMS320 DSP platform in the early 1990s. Our application then was wideband satellite downlinks, where we were limited to the amount of threshold extension that was possible due to the wide Carson's Rule bandwidth. In the present application we might be able to do better, but TEDs aren't my first choice because I'd like to avoid that front-end FM threshold effect.

In principle, direct detection of the Bessel components at RF should work better IF the analysis is tractable. Hence my original question about how distinct those sidebands are in the presence of typical program channel audio.

Grant,

I think we talked about this before. I will give you a simple example that demonstrats the problem you need to solve. If the FM station is sending only the 19 kHz pilot and nothing else then you have the 2 sidebands serarated from the carrier by 19 kHz as you estimated. At this point, you could use a narrow band receiver and detect or do whatever you want with those two sidebands. OK so far so good.

But now consider the FM station is also xmitting a bass drum hit say

100 Hz. The main carrier AND THE 19 kHz SIDEBANDS are ALL now going to sweep back and forth at 100 Hz by up to +/- 75 kHz. So if you want to use a narrow band device to detect the 19 kHz sidebands, you need to somehow deal with (track them) the fact that they are sweeping around with the carrier.

I'm not saying it can't be done, but that is the issue you have to solve.

As I told you last time, you want look up the literature about "threshold extension" demodulators.

What is the application that you need to detect the 19kHz pilot tone if the main carrier is so far below threshold?

Mark

The 19kHz pilot appears as the periodicity in the FM spectrum.

1. Take the energy spectrum of the FM signal.
2. Calculate the parameter: X = sum { W(F) + W(F+19kHz) } for all frequencies.
3. Compare X against the total energy of the signal.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Oops, got it wrong. Instead of (2) run a 19kHz Goertzel along the spectrum, so it will pick up all 19kHz periodicities.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

A lot of us don't get a.b.s.e., or any of the alt groups.

Don't have a website where you can post pix? I've notice that you have an ftp server where you post stuff occasionally - if you let them be served up by your http server, the links would be clickable.

If you don't know how to do this, I'd be happy to do it for you, for a minor consideration.

Cheers! RIch