It's probably RDS or some variant of it. RDS gets double-sideband modulated on a suppressed 57 kHz carrier, that gets added to the FM multiplex signal. In the multiplex signal, it occupies the frequency band from 2375 Hz below to 2375 Hz above 57 kHz, with a notch in its spectrum at 57 kHz.
The peak frequency deviation due to only the RDS signal is between 1 kHz and 7.5 kHz. The most usual value is 2 kHz.
In the case of stereo broadcasts, the 57 kHz subcarrier is phase locked to the 19 kHz pilot tone to minimize interference.
The peak frequency deviation due to only the RDS signal is between 1 kHz and 7.5 kHz. The most usual value is 2 kHz.
In the case of stereo broadcasts, the 57 kHz subcarrier is phase locked to the 19 kHz pilot tone to minimize interference.
Last edited:
Supposedly this detector doesn't work at the resonant peak, but is rather an edge detector.
Best regards!
Best regards!
I don't know the answer, but if I had to guess I'd say it is some transformerless variant of a Foster-Seeley detector. At least there are two rectifying diodes the rectified signals of which get subtracted and that apparently get IF signals with a frequency-dependent phase difference. There is no large capacitor that suppresses AM, as in a ratio detector.
Last edited:
Thanks for all replies. Yesterday I got 3°dosis of Moderna vaccine and feel very tyred. During week I'll start working in the IF strip. My idea is a two stages of EF184 pentodes, and the IF trafos made of my own.
I read some old papers about reckoning mutual inductance M and coupling coefficient k. As I haven't Q meter, the idea is to make two series of calculus, one estimating a Q of say, 40 and another of 100 and finally adjust coil separation to get the desired bandpass action. This will gave me some extreme values and I suppose that the final value will be between them. I saved some pieces of plastic tubes to make the bobbins for the traffos and use again, copper tubing inside the bobbin to adjust them at resonance. Coils supposedly will be more or less square, say same diammeter as long as it is said this geometry gives better Q. Tunning capacitance will be 30pf silvered mica brought for this project.
Perhaps I use the above scheme to get the AFC signal indepently of the audio detector as it uses few more materials. Obviouly diodes will be 6AL5 type as I want an all-tube state of art in my project.
But again, I feel very tyred now.
Enjoy posting FM topics here 🙂.
I read some old papers about reckoning mutual inductance M and coupling coefficient k. As I haven't Q meter, the idea is to make two series of calculus, one estimating a Q of say, 40 and another of 100 and finally adjust coil separation to get the desired bandpass action. This will gave me some extreme values and I suppose that the final value will be between them. I saved some pieces of plastic tubes to make the bobbins for the traffos and use again, copper tubing inside the bobbin to adjust them at resonance. Coils supposedly will be more or less square, say same diammeter as long as it is said this geometry gives better Q. Tunning capacitance will be 30pf silvered mica brought for this project.
Perhaps I use the above scheme to get the AFC signal indepently of the audio detector as it uses few more materials. Obviouly diodes will be 6AL5 type as I want an all-tube state of art in my project.
But again, I feel very tyred now.
Enjoy posting FM topics here 🙂.
Last edited:
Long ago at Elektor I designed a simple 3 transistor FM receiver with such a discriminator. As that was before the internet era I didn't bother to look for names. In the circuit diagram L7 & C14 is a series resonance and at resonance, carry equal voltages. At higher frequencies the voltage over L7 is higher than that over C14 and below resonance it's the reverse. AFC can be derived from it, frequency-feedback as well so despite the low IF it could even be used for stereo.
Attachments
Very intetesting. My pic is from a mazine of my father's era, say, 70's. But, has it a name from his creator? and how to calculate (or estimate) the limits of linearity or BW?, I suppose relationed to LC values and load resistors IMHO.
The principle is clear for me from the very start because it is explained in such an artcle from I took the pic, but my suspicion is that diodes in parallel to tank components, load them severely.
Why a varactor in series with oscillator gate?
The principle is clear for me from the very start because it is explained in such an artcle from I took the pic, but my suspicion is that diodes in parallel to tank components, load them severely.
Why a varactor in series with oscillator gate?
I don't know who created the discriminator based on series-resonance but at 1st sight it looks like a logical choice, also because no transformer is needed. But linearity can be a problem, because for proper functioning the series LC needs a constant voltage. In the circuit this was more or less fixed by the limiter diodes D2 and D3 plus the IF mosfet running at ~20 mA. As the IF is ~200 KHz it's obvious that for 75 KHz deviation heavy loading even is required. Hence also the linearity improvement from frequency feedback. When running the discriminator on a high frequency (your choice remains 12 MHz?) diode loading might need to be minimized. The easy way would be to use a lower value of the coil and a higher value of the cap. The source impedance then has to be lowered as well.
The varactor in series with the gate served the purpose of stabilizing the oscillator. A higher positive voltage on g2 increases the RF gain via g1 but decreases the gain via g2, assisted by the decreasing capacity of the varactor. That was something you couldn't learn from a data book.
It looks very like the detector described in E.H. Armstrong, Frequency Modulation - Noise Reduction, Proc. IRE vol. 24 no 5, May 1936, Figure 7.
Thanks again sirs.
Yes, my idea is to take the IF value of 12MHz. But I am able to change it if some good rasons to change it are explained to me. Definitively, the difference with the usual 10.7MHz is about 10% and I don't suspect that something is easy to build at 10.7 than can't be done at 12MHz.
Again, I haven't modern instruments than my restored EICO 368, a RLC digital meter and a 40MHz oscillloscope. No Q meter.
I'll search for the literature suggested above.
Very cloudy and wet today here. 100% humidity and very dense and dark is the weather today.
Yes, my idea is to take the IF value of 12MHz. But I am able to change it if some good rasons to change it are explained to me. Definitively, the difference with the usual 10.7MHz is about 10% and I don't suspect that something is easy to build at 10.7 than can't be done at 12MHz.
Again, I haven't modern instruments than my restored EICO 368, a RLC digital meter and a 40MHz oscillloscope. No Q meter.
I'll search for the literature suggested above.
Very cloudy and wet today here. 100% humidity and very dense and dark is the weather today.
The main reason to change it is that you will never find an FM generator with a 12MHz marker to align it with. Use 10.7MHz like nearly everyone else.
A higher than 10.7 MHz IF could be the better choice because when high enough, a fixed filter handles the input selectivity for image rejection so only oscillator tuning is needed. The disadvantage is the possibility of IF harmonics leaking to the input but can be prevented easily with a 2nd conversion. Thus, the front end of a high end FM tuner with tubes could consist of RF amp 7784, 87-108 BPF, SBM with ECC88 and ECC88 LO plus buffer.
My EICO generator had a 12MHz xtal to adjust the center of the IF strip. So it will not be a problem. With regard to harmonics, eventually I shall take the corrective measures to reduce them to a minimum. I had the idea of using a low Q trap in the lower triode of the cascode cathode. But let's see if it is really needed.
Once again, thanks for your comments and patience to me.
Once again, thanks for your comments and patience to me.
Attachments
Re-thinking your statement, I found very doubtful may occur. Maths says:
12MHz × 8° harmonic = 96MHz.
12MHz × 9°harmonic = 108MHz.
So 8th harmonic is out of band, although close. 9th is just in the border.
Let me see what happens as the project maturates.
I remember the issue from my tube FM tuners: 9X10.7 was quite a nuisance (FM band back then in Western Europe, only to 104 MHz. Compared to semiconductor designs, IF voltage of a limiter 10X higher or more, plus longer lines carrying the high IF voltage. For the 1st IF, 28 MHz would be a good choice, allowing fixed input filter. So it will be interesting to see how your project develops. My "home" projects were and are designed for performance because at the job, standard performance and simplicity always had priority.
Ok. But my material and instruments are limited. To import new tubes is almost impossible because local politics and regulations. So I need to use what I could save years ago.
Can you imagine a good design using compactrons? I have several of them new and with litle use. There is a dual pentode with gms of 30mS with 10mA +/-. More than double than EF184, but the stabilty may be troublesome.
Do you also want to dialog in Spanish by whatsapp?. Send me an email to lw1dseATgmailDOTcom or lw1dseATyahooDOTcom. I send my telephone #.
Kind regards.
Can you imagine a good design using compactrons? I have several of them new and with litle use. There is a dual pentode with gms of 30mS with 10mA +/-. More than double than EF184, but the stabilty may be troublesome.
Do you also want to dialog in Spanish by whatsapp?. Send me an email to lw1dseATgmailDOTcom or lw1dseATyahooDOTcom. I send my telephone #.
Kind regards.
I bought several matched pairs of the audio-famous 6E5P and as the tetrode was designed for HF, it's likely it can be used successfully at VHF too. It also has Gm~30 mA/V, Req = 350 (!) so much lower when used as triode. With the compactrons that might be similar, provide the list and I'll take a look. Stability won't be a problem. My knowledge of Spanish is limited to the issue "local politics and regulations" ;-) and my "science & engineering" range of discussion is limited to Dutch, English and German - French would have to be refreshed, never used during study / work.
I still have a Facebook account though and that provides fairly good translation. Same user name, too many Europeans are named "Jan". Living very rurally has many advantages, but fast internet (for video conferencing etc.) isn't one of them. I wouldn't mind losing "friends / followers" who don't like to see Facebook discussions on tuners (AM, FM), HiFi amplifiers etc.
I still have a Facebook account though and that provides fairly good translation. Same user name, too many Europeans are named "Jan". Living very rurally has many advantages, but fast internet (for video conferencing etc.) isn't one of them. I wouldn't mind losing "friends / followers" who don't like to see Facebook discussions on tuners (AM, FM), HiFi amplifiers etc.
Ok. Forget then. Argentine politics are very stupid and elementary when taking economical measures and we (the common population) suffer its consequences.
I have several unknowns and doubts but during the experiments it is highly possibly that they be disipated. I found valuable info and data in old papers and magazines and also, many ideas from Google Patents.
But one I still is in doubt is the limiter stage. It is clear that it roulgly operates as a class C amplifier more than a class A one. It is said that low screen voltage is needed to secure saturation of plate current and safe operation under no singnal condition and no grid leak bias. But why is needed a low plate voltage too if pentode current is cuasi independent of voltage at anode and with normal anode voltage, more signal can reach detector stage?
Many thanks one more once.
I have several unknowns and doubts but during the experiments it is highly possibly that they be disipated. I found valuable info and data in old papers and magazines and also, many ideas from Google Patents.
But one I still is in doubt is the limiter stage. It is clear that it roulgly operates as a class C amplifier more than a class A one. It is said that low screen voltage is needed to secure saturation of plate current and safe operation under no singnal condition and no grid leak bias. But why is needed a low plate voltage too if pentode current is cuasi independent of voltage at anode and with normal anode voltage, more signal can reach detector stage?
Many thanks one more once.