Foster-Seely FM Discriminators

[Jim - G4MEZ]

I'm the proud owner of a Pye FenManII A beautiful set which puts my Bush VHF64 somewhat in the shade performance-wise. It is not yet fully restored and much still needs to be done.

I'm chasing a strange problem in the FM discriminator (see other post) and I see the Pye uses a Foster-Seely type whilst most contemporary sets use a Ratio Detector.

Why did Pye go its own way here and not use the more common Ratio Detector? Is there some performance advantage? I cannot find a satisfactory answer in any of my text books.

Put me out of my misery please

[ukcol]

As I understand it the Foster-Seely detector has a better performance than the ratio detector from a linearity point of view. Or put another way it produces less distortion. It's dis-advantage is that it requires a limiter stage before it to get rid of all the AM that may be on the carrier.

The ratio detector is not sensitive to AM on the carrier and so for general use is preferred,

No doubt Pye were happy to go to the expense of an extra limiting IF stage and use the better detector for this up market set.

I use a Fenman II as a daily use set and very nice it is too.

[Skywave]

My understanding goes like this:

Unlike the Ratio Detector (R.D.), the Foster-Seeley (F.S.) is not self-limiting with regard to amplitude changes in its input. Therefore, with a F.S., a stage of amplitude-limiting is necessary, prior to the detector.
The F.S. produces less distortion than the R.D. (F.S. up to 1% compared to 3% for the R.D., both for a 75 kHz deviation).
The F.S. produces approx. twice as much output signal compared to the R.D.
The R.D produces an AGC voltage which can (and should) be used to control the gain of amplifying stages prior to the detector. It can also be used as a tuning indicator. This voltage is not available from the F.S.

Generally, the R.D. is cheaper to implement than the F.S. in a given design. That's probably the main reason why it was almost universally used in FM broadcast receivers (although these days, PLL's are the method often used).

Furthermore, I believe that the R.D. was derived from the F.S. - but I may be wrong on this

Al / Skywave.

[oldeurope]

Hi,
I don't agree here.
Foster Seely tuned circuits are used in the discriminator and in the ratio.
There is a limiter arrangement called "dynamic limiter".
The ratio is a clever combination of both dynamic limiter and discriminator.
Thus if you want to compare them, you must campare a discriminator
with dynamic limiter added to a ratio. There is no reason why one of these
should give more or less distortion.
To use the ratio arrangement you needed a licence (from RCA?).
That is all.

Kind regards,
Darius

[Brian R Pateman]

Originally Posted by Jim - G4MEZ

Why did Pye go its own way here and not use the more common Ratio Detector? Is there some performance advantage?

The Fenman II was state of the art when it was released and an expensive set. As other posters suggest, Pye thought it was worth the additional cost of providing limiting on their flagship offering to provide the best possible performance.

(Regulars on here will know of my bias in favour of these excellent sets!)

Regards,

[ukcol]

Originally Posted by Skywave

Furthermore, I believe that the R.D. was derived from the F.S. - but I may be wrong on this

Al / Skywave.

I'm pretty sure you are right about that Al.

I have been looking through my books for a definitive answer because I am sure I've read it quite recently but I haven't been able to find the text.

One interesting bit of information that I did find was in Chas Miller's book Valve Radio & Audio Repair Handbook, and was this. The Foster Seeley Discriminator was invented by D.E.Foster and S.W.Seeley in 1936 and the subject of a paper read to the Institute of Radio Engineers in March 1938.

This of course predates its use as an FM detector in domestic receivers. Chas's book describes its use in generating the error voltage in AFC systems.

[Skywave]

Originally Posted by ukcol

I'm pretty sure you are right about that Al.

I have been looking through my books for a definitive answer because I am sure I've read it quite recently but I haven't been able to find the text.

OK - well just for the record, the two main sources I used were:

(a) "Electronic Communications Systems", 2nd. Ed., by George Kennedy (McGraw Hill), 11th. printing 1983;

(b) "Radio & Television Engineers' Reference Book", Editor J.P. Hawker (aka G3VA), 4th. Ed., 1963.

Various ARRL handbooks (and other USA-origin reference books) and assorted publications of the RSGB do make some useful references, but of course these are tailored to AmRad use of FM = NBFM in the AmRad world.

Unfortunately, my copy of "Langford" makes no reference to FM - probably because it's a very old issue. (Ditto "Foundations", MGS). However, I am reliably informed that improved forms of discriminator are referred to by Langford in a 1960 publication and Sturley in 1965.

Al / Skywave

[Hybrid tellies]

Originally Posted by Skywave

My understanding goes like this:

Unlike the Ratio Detector (R.D.), the Foster-Seeley (F.S.) is not self-limiting with regard to amplitude changes in its input. Therefore, with a F.S., a stage of amplitude-limiting is necessary, prior to the detector.
The F.S. produces less distortion than the R.D. (F.S. up to 1% compared to 3% for the R.D., both for a 75 kHz deviation).
The F.S. produces approx. twice as much output signal compared to the R.D.
The R.D produces an AGC voltage which can (and should) be used to control the gain of amplifying stages prior to the detector. It can also be used as a tuning indicator. This voltage is not available from the F.S.

Generally, the R.D. is cheaper to implement than the F.S. in a given design. That's probably the main reason why it was almost universally used in FM broadcast receivers (although these days, PLL's are the method often used).

Furthermore, I believe that the R.D. was derived from the F.S. - but I may be wrong on this

Al / Skywave.

Just read this up in the excellant FM Servicing handbook (2nd edition) by Gordon J King which matches up almost word for word with Skywave's description. An interesting point, Gordon King says that the Foster-Seeley circuit is often refered to as the Phase Discriminator.
The earliest FM Detector was developed back in the 1920's and was called the Travis Frequency convertor which is explained in the book to some detail before moving on to the Phase Discriminator and then the Ratio Detector. Apparently the development of FM dates back to the early 1920's.

[dazzlevision]

Hello,

Don't all FM radio receivers (and 6MHz intercarrier sound in TVs) use a limiter, prior to the FM demodulator - of whatever type? The limiter is incorporated in the final IF amp stage of a radio (AM/FM or FM only sets), but with bias adjusted to clip the IF signal - on FM only.

I'm pretty sure that a practical method of Frequency Modulation - for radio - was developed/invented by Mr Edwin Armstrong (USA) in 1933.

Edwin Armstrong received U.S. patent 1,342,885 for a "Method of Receiving High-Frequency Oscillations Radio" for his FM technology.

He also invented regeneration (1913) and superheterodyning (1910s).

Just "google" for more information!

Regards,

Dazzlevision

[Skywave]

Originally Posted by dazzlevision

Edwin Armstrong . . . . also invented regeneration (1913) and superheterodyning (1910s).
Dazzlevision

Mr. Armstrong is also credited with discovering the super-regenerative principle. AKAIK the story goes like this . . . .

Apparently, it all arose accidentally and quite spontaneously whilst he was doing some experimental work with "ordinary" regenerative receivers. Occaisionally, various stations could be received that with their known ERP and distance, couldn't readily be accounted for with the cct. he was using. The fact that the effect seemed to "come and go" at random made an investigation of the cause difficult to pin down - to put it mildly. It was only several months later - and after a lot of work - that the "penny dropped" - and the super-regenerative receiver was born.

Al / Skywave