What causes the capture effect in FM modulation?

[FrankJ777]

My understanding of the FM capture effect is that if an FM receiver receives two, on frequency signals, it suppresses the weaker one and demodulates the stronger one, given that the difference in signal strength is significant enough. This differs from the case of AM where the two signals are summed together at the receiver. I don't understand how the capture effect work. I've tried working it out mathematically, and in block diagram form, but I still don't understand the phenomenon.

Can anyone explain what causes the capture effect?

Also I've read that the capture effect occurs with all angular modulated signals; i.e. FM, PM, PSK, QPSK, QAM, etc. From what I remember from my communications course, when demodulating PSK, QPSK, and QAM, the receiver decomposes the signals into an I and Q signal, and they are individually demodulated using AM like techniques. So, is it accurate that the capture effect is present with these types of modulation schemes?

Thanks!

 

[meBigGuy]

Capture is caused by limiting the signal before FM demodulation. A side effect of limiting is that the weaker signal has little effect of the instantaneous frequency, so is not heard.

http://www.dtic.mil/dtic/tr/fulltext/u2/a208479.pdf is an analysis that I am posting blind.

 

[sophiecentaur]

One way of explaining the effect is to think in terms of the effect of summing the two signals on the instantaneous phase of the received signal. If the 'wanted' signal has a significantly greater amplitude than the unwanted signal then the receiver will ' see' a carrier vector, which is rotating back and forth (as the frequency is modulated), to which is added a smaller carrier vector. If this interfering vector is smaller than the wanted vector then it will appear as a 'fuzzy end' on the wanted vector, never taking the resultant around the origin, so the wanted vector will be seen to rotate about the origin in step with its modulating signal. The interfering vector will only ever appear as a small perturbation on the phase of the wanted vector (perhaps a matter of 90 degrees or less). For high deviation FM, the demodulated signal will be totally dominated by the modulation of the wanted signal ( several multiples of 360 degrees of phase swing).
This effect also results in an improvement in demodulated signal to noise ratio (the FM noise advantage) of 20dB or so, in the case of regular domestic broadcasts.

NB Wide deviation also calls for a wider bandwidth, of course. This admits more channel noise and there is a level of Carrier to Noise ratio, below which the peaks of the added noise will exceed the carrier power. Below this 'Threshold', the phase of the RF + noise can flip around the origin, producing an impulsive noise spike on the demodulated signal. The audible noise suddenly gets much worse. The signal becomes unusable all of a sudden and the channel 'crashes'.

Radio communication channels often use low deviation FM, which performs much the same as AM, because the Improvement and Threshold effects are small. But, as with AM, the lower quality signal 'holds up' much better than the broadcast quality signal when reception is poor.

 

[FrankJ777]

Thanks guys. I think both our explanations help, I didn't notice anything about the limiter causing FM capture in the paper though. It seems like the author attributes it to the low pass filtering process in the FM receiver.

The other part of my question was; does or how does thecapture effect happen with phase modulation, such as QPSK or BPSK, which from what I understand are demodulated with techniques much like AM once the I and Q components are separated.

 

[sophiecentaur]

The equivalent 'mod index' for psk is low so the rf bandwidth is comparable with the signal bandwidth, I wouldn't have expected any advantage / capture. The amplitude and phase of wanted and unwanted signal would contribute to a resultant vector on equal terms, in principle. But the coding and synchronisation of the wanted signal can make a big difference to the error rate (constituting an advantage of another kind).

 

[sophiecentaur]

Thanks guys. I think both our explanations help, I didn't notice anything about the limiter causing FM capture in the paper though. It seems like the author attributes it to the low pass filtering process in the FM receiver.

The other part of my question was; does or how does thecapture effect happen with phase modulation, such as QPSK or BPSK, which from what I understand are demodulated with techniques much like AM once the I and Q components are separated.

That's because there are demodulators that do not use limiting. The limiting type of discriminator is very convenient design but it is not 'essential' for FM discrimination. The "ratio detector' circuit is one of those, iirc. The level is maintained by a straightforward AGC loop.