FM Stereo Spectrum question

[Apfelmus]

Hi all,

having repaired the power suppy of my R&S transmitter I wanted to do a full realignment. Without a modulation analyzer and stereo decoder for doing some of the stuff in the manual, I wanted to use a spectrum analyser looking at the RF signal.

See pic#1 for a scholar FM stereo spectrum on Wikipedia, showing the bandshape for L+R, the pilot, L-R around 38kHz and some more signals for RDS. So with L=R there should be nothing in the 38kHz region. And with L =-R nothing should be on the carrier and all around 38kHz.

Here is what I get from a modulated 1kHz L+R signal (pic#2). It spreads from -75kHz to +75kHz! Which is as it should be, according to the settings. But it smears all the way through the areas where the pilot and the L-R signal is supposed to be!

Pic#3 is pilot tone alone. Very present at 38khz and even further on. Pic#4 is with modulation L+R added. Pic#5 with L only.

I couldn't figure out how to make use of these waveforms. How does the scholar spectrum from pic#1 relate with i.e. pic#5?

cheers
Martin

[Dickie]

I think you are confusing the modulation spectrum (pic 1) with the RF spectrum (pics2-5).

[Radio Wrangler]

The first picture is the spectrum of the baseband (zenith pilot tone stereo audio) which is put into an FM transmitter

The other pictures are all spectra of the OUTPUT of an FM transmitter.

If you used an FM demodulator feeding a low frequency spectrum analyser then you would start to see pictures similar to your first one, though what terms would be present would depend on the audio you were transmitting and whether you were incoluding RDS etc.

In simple terms, the first picture is the spectrum that would be found at the terminals of the speaker of a radio, the other four are the sspectrum as it would appear at the antenna connector.

Your spectrum analyser has not been set to a narrow enough resolution bandwidth to separate all the FM components created by such complicated modulation.

My portable spectrum analyser can be made to stop scanning and to demodulate AM or FM signals with the resolution bandwidth set wide enough to grab a whole FM (orAM) signal, and the digital section can take an FFT to give me the spectrum of the demodulated signal. Some analysers can do what you seem to want to do.

David

[Apfelmus]

Thanks! Must have been put off track by the term "baseband", thought of it as the unmodulated carrier.

Wanted to use the transmitter output as I had no other signal to look at. So it requires some FM receiver and demodulator, preferrably with an MPX output to see the higher frequencies.

If I understand right, using a white noise signal would still require this setup.

cheers
Martin

[Radio Wrangler]

The audio output from the MPX decoder of an FM tuner will usually not contain the higher frequencies. They will usually be removed by filtering in the decoder. If they were allowed to go out, they would cause unwanted tweak and whistle noises by beating with the bias oscillator of any tape recorder used on them. So "MPX filters" are common at the end of an FM tuner.

A mono output or tuner might give you them, but it depends on what audio filtering the designer chose, and often the IF bandwidth is too narrow for the FM sidebands of those higher frequencies to be received.

David

[Julesomega]

Access to the MPX decoder will allow you to look up the IC data and find the MPX output before the filtering

[Apfelmus]

One experiment I made is to hook the signal up to an HP MDA (Modulation Domain Analyser, HP53310). It works as far as I can see the modulation, but there is no output of it anywhere on the analyser. The signal is probably all in the digital realm, so I cannot even probe it somewhere in the unit.

Must check my tuner "collection" for MPX outputs. There might be some older tube models readied for stereo with an external decoder.

cheers
Martin