FM Capture Ratio

Forum FM enthusiasts, I am looking for good refrences that analyse and explain how capture ratio works because of FM limiting before demodulation.

I have read these articles by RM contributer Prof Ing Dietmar Rudolph, with help from Google-translate

http://www.radiomuseum.org//forumdata/upload/FM_Demodulation_WS0506.pdf

http://www.radiomuseum.org//forumdata/upload/FM_Demodulatoren.pdf

http://www.radiomuseum.org//forumdata/upload/Frequenz-Demodulation_72.pdf

These references explain the usefulness of capture ratio and it's effect on FM reception.

In particular, I am trying to understand what is lost if FM demodulation is done without limiting in amplification stages, or without limiting at the detector, as is the case with a flank detector. But I am also interested in the fundamentals of the capture ratio mechanism.

The best intuitive understanding I can manage, is that when limiting occurs, more of the weaker signal is lost than the stronger signal. When limiting is very strong, only zero crossings of the signal survive, and these are strongly dominated by the stronger signal.

With thanks in advance for any help,

-Joe

To thank the Author because you find the post helpful or well done.

 

To thank the Author because you find the post helpful or well done.

Hello Hans,

Thanks for the reference to Recklinghausen. After following your link, I realized that he was a major force in the early days of HI-FI in the USA, and in particular where I live. He set up the first FM stereo transmissions at WCRB, which is still on the air with a program of popular classical music.

One of the tuners in my home lab is a H H Scott 314 wideband mono tuner. This tuner drives one of the famous single pentode Grundig FM stereo decoders, with very good stereo output sound.

Since I posted the request for material on capture effect, I found this IEEE article from 1976 with a good bibliography from the 1950's:

The Capture Effect in FM Receivers by KRIJN LEENTVAAR AND JAN H. FLINT - IEEE Transaction on communications Vol Com-24, No 5 , MAY 1976. This article is available from the IEEE for a $10 fee from IEEE members. But the bibliography points to several freely available articles from the 1950's at MIT.

[1] J. W. Alexander, “Een eenvoudige rekenwijze voor het berekenen van stroomkringen waar in frequentie-gemoduleerde spanningen werken,” Tijdschrift van het Nederlands Radiogenootschap (in
Dutch), vol. XI, 1946.

[2] E. J. Baghdady, “Interference rejection in F.M. receivers,” Electron. Res. Lab., M.I.T., Cambridge, MA, Tech. Rep. 252, Sept.1956.  free download at http://dspace.mit.edu/handle/1721.1/4817 

[3] M. S. Corrington, “Frequency modulation distortion caused by common-and adjacent-channel interference,” RCA Review,Dec. 1946.

[4] J. Granlund, “Interference in frequency-modulation reception,” Tech. report 42, Electron. Res. Lab., M.I.T., Cambridge, MA, Jan. 1949.  free download at http://dspace.mit.edu/handle/1721.1/5006 

[5] P. Giittinger, “Die Gegenseitige Beeinflussing zweier frcquenzmodulierter Wellen in Amplitude-Begrenzer,” Brown Boveri Mitteilungen, Sept. 1944.

[6] F. L. H. M. Stumpers, “Interference problems in frequency modulation,”Philips Res. Rep., vol. 2, Apr. 1947.

However, I would still like to hear any practical appreciation of the capture effect or it's benefits to FM reception.

Regards,

-Joe

To thank the Author because you find the post helpful or well done.

Those MIT papers are math-heavy, not intuitive.

> what is lost if FM demodulation is done without limiting

Most simple "FM demodulators" are AM demodulators with a slope.

If you do not limit, all AM interference is detected. In particular, impulse (lightning, engine spark hash) stronger than desired signal comes right through.

With limiting, hash is no greater than desired signal, average frequency is nowhere near tuned frequency, the tuned-circuit will fly-wheel to the desired signal and tend to reject random impulse hash.

Imagine a child's merry-go-round. Daddy gives it a push once per turn. If a bully is allowed to smack it with a heavy hammer, rotation may be badly upset. If bully is limited to no more force than Daddy's hand, and smacks randomly, rotation will be pretty much as Daddy intends.

> practical appreciation of the capture effect or its benefits to FM reception.

As you say: with a sufficiently limited signal, the tuned circuit will be dominated by the strongest signal, and will reject other carriers.

If you drive from NYC to Philadelphia you can hear it. I forget what frequency, but two major FM stations (one near each city) share a frequency. Between Trenton and Delanco, your radio "shifts" from one city to the other, several times, as each station becomes "stronger". (In a flat perfect world, there would be one shift; hills and roadside wires give several shifts.) While there may be a few instants of "both" stations, mostly you get one or the other, not both.

In dense urban areas, at a backed-up stop-sign, you occasionally find two stations alternate as you creep up to the intersection-- local reflections double or null each signal at different places.

This in contrast to AM, where any simple detector will receive two (or more) stations at once in proportion to strength. I have listened to CFRB on top of WCBS many rainy nights. Both are highly compressed, so when one is 10dB down the other dominates your ear, but the weaker signal is audible especialy when the louder one "takes a breath". In this case WCBS was long-local or first-skip, CFRB was second or third skip, so as I drove along CFRB faded in and out every mile. WCBS ~~40 miles over Coppermine Hill, CFRB 500 miles over Lake Ontario and NY/PA hills.

When locating AM stations co-channel, large geographic gaps must be allowed because the in-between signal is "bad" for both stations. On FM, aside from line-of-sight limit, stations may be co-channel and close because in the overlap zone, one or the other but not both will dominate any specific fixed receiver location.

To thank the Author because you find the post helpful or well done.

In a FM wave information lies in its momentary frequency and not in its amplitude. Therefore any unwanted amplitude modulation in a received FM will give a distortion of the demodulated signal, when demodulated with an indirect FM demodulator which first changes frequency deviation to amplitude devialtion, e.g. ratio detector or discriminator.

So it is necessary to use an amplitude limiter to get rid of disturbing amplitude fluctuations resulting from disturbance in the radio channel, like noise, impulses and so on.

On the other hand, the momentary frequency of an FM signal is uniquely defined by the zero crossings of the FM signal, and so is the demodulated audio signal. Regarding the sampling theorem, the samples are much more often than the upper corner frequency of the audio signal necessitates. This is somewhat similar to a one bit D/A converter in a CD player, which also has to use oversampling. 

Now have a physical look to the problem. If two sinusiodal oscillations with nearly the same frequency, but different amplitudes are added, some kind of an oscillation nearly like AM will result. However, if you have a look to the zero crossings of this resulting oscillation, no major difference to the zero crossings of the sinusoidal oscillation with the bigger amplitude will be observed. The reason for this is the steepness of the time function at this points, and adding something with moderate amplitude will not change much the point of the zero crossing. This effect also holds for FM modulated signals with different amplitudes. The signal with the major a amplitude will dominate, and so the audio signal modulated on it. That is, the stronger signal will be captured. But this will only happen, if a pefect limiter is applied.

The capture ratio is defined for a suppression of the modulation of the weaker FM signal by some 30 dB. Excellent limiters provide a capture ratio of about 1/2 dB.

Thererfore, at the limit of the supply area of a TX, jumping from one program to another of a different TX on the same frequency is normal. This is due to the fading from which the reception suffers according to multipath propagation.

Usually in urban environmet, many reflections arrive at the receiver. With a FM signal these refections produce a hissing sound. Sometimes, this can be heard when a car stops at the light. Or it will be observed with a small radio in the bathroom. In any case, this is an indication for an insufficient limiter.

A limiter not only will free FM reception from noise due to amplitude disturbances, but also from echo disturbance. (Please recall that propagation delay is mayor problem with digital transmissions.) 

Regards,

Dietmar

To thank the Author because you find the post helpful or well done.