Jerry Avins <jya@ieee.org> wrote:>The FM detector produces -- before deemphasis -- baseband audio (the sum >channel), a difference signal modulated onto a nearly-40-MHz subcarrier,38 KHz, not MHz :)>and a half-carrier-frequency pilot tone. Those three components areYes, 19 KHz.>decoded into left and right audio channels. How the FM was demodulated >to get them makes no difference at all.ACK.>JerryRalph.
FM Demodulation
Started by ●January 17, 2007
Reply by ●January 18, 20072007-01-18
Reply by ●January 18, 20072007-01-18
Eric Jacobsen <eric.jacobsen@ieee.org> wrote:>As has been stated, the vast majority of analog FM receivers for >broadcast use PLLs, mostly because it's a tiny circuit with just a few >transistors compared to other (especially digital) implementations and >it does work quite well.They are not used for demodulation, but as 1st LO. The normal demod for analogue FM apps is a discriminator.
Reply by ●January 18, 20072007-01-18
Eric Jacobsen <eric.jacobsen@ieee.org> writes:> On Wed, 17 Jan 2007 23:00:39 -0500, Randy Yates <yates@ieee.org> > wrote: > >>sampson164@gmail.com writes: >> >>> Randy Yates wrote: >>>> Gentle and Wise Readers of comp.dsp, >>>> >>>> Two ways to demodulate FM are: >>>> >>>> 1. Find the instantaneous phase (e.g., by examining the phase of the >>>> analytic signal) and compute the phase differences, i.e, f = dtheta/dt. >>>> >>>> 2. Lock a PLL to the FM signal and use the VCO control voltage as the >>>> demodulated signal. >>>> >>>> It seems to me that method 1 is much simpler. Why would anyone do FM >>>> demodulation via method 2? >>>> >>>> Let me state up front that I have an idea, but I'd rather get others' >>>> ideas first without biasing them. >>>> -- >>>> % Randy Yates % "She tells me that she likes me very much, >>>> %% Fuquay-Varina, NC % but when I try to touch, she makes it >>>> %%% 919-577-9882 % all too clear." >>>> %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO >>>> http://home.earthlink.net/~yatescr >>> >>> The SNR threshold is lower with a PLL. Search for "Threshold Extension" >>> for details. >> >>That's it, John - that's what I was looking for. Apparently it's >>fairly well-known - I found it in [couch] after my original post It's >>apparently also known as FMFB (FM with feedback), which is referenced >>in both [couch] and [schwartcommtecniques]. >> >>Does anyone know how prevalent the PLL technique is in current >>demods (TVs, FM broadcast, etc.)? Is it pretty much standard >>fare now or are they somewhat unique? > > As has been stated, the vast majority of analog FM receivers for > broadcast use PLLs, mostly because it's a tiny circuit with just a few > transistors compared to other (especially digital) implementations and > it does work quite well. > > So the cost/performance tradeoff is hard to beat, and the power > consumption is tiny.Sure - that makes sense. I guess I was looking at it from a S/W POV. Also, if this helps improve performance for analog FM, wouldn't it do the same for m-ary FSKs and variants? Would GMSK be improved with this type of receiver? -- % Randy Yates % "And all that I can do %% Fuquay-Varina, NC % is say I'm sorry, %%% 919-577-9882 % that's the way it goes..." %%%% <yates@ieee.org> % Getting To The Point', *Balance of Power*, ELO http://home.earthlink.net/~yatescr
Reply by ●January 18, 20072007-01-18
Ralph A. Schmid, DK5RAS wrote:> Jerry Avins <jya@ieee.org> wrote: > >> The FM detector produces -- before deemphasis -- baseband audio (the sum >> channel), a difference signal modulated onto a nearly-40-MHz subcarrier, > > 38 KHz, not MHz :)Thanks. I really knew that, honest!>> and a half-carrier-frequency pilot tone. Those three components are > > Yes, 19 KHz. > >> decoded into left and right audio channels. How the FM was demodulated >> to get them makes no difference at all. > > ACK.Que? Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●January 19, 20072007-01-19
"Randy Yates" <yates@ieee.org> wrote in message news:m3wt3ltfzc.fsf@ieee.org...> Gentle and Wise Readers of comp.dsp, > > Two ways to demodulate FM are: > > 1. Find the instantaneous phase (e.g., by examining the phase of the > analytic signal) and compute the phase differences, i.e, f = dtheta/dt. > > 2. Lock a PLL to the FM signal and use the VCO control voltage as the > demodulated signal. > > It seems to me that method 1 is much simpler. Why would anyone do FM > demodulation via method 2? > > Let me state up front that I have an idea, but I'd rather get others' > ideas first without biasing them. > -- > % Randy Yates % "She tells me that she likes me verymuch,> %% Fuquay-Varina, NC % but when I try to touch, she makesit> %%% 919-577-9882 % all tooclear."> %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO > http://home.earthlink.net/~yatescrFor 2 it's the filter output, not the VCO output that is the demodulated FM. A PLL will gave an improvement of around 3dB in demodulated SNR over other methods. F. -- Posted via a free Usenet account from http://www.teranews.com
Reply by ●January 19, 20072007-01-19
"Ralph A. Schmid, DK5RAS" <dk5ras@dk5ras.de> wrote in message news:t4gvq29m2gn8rq41j1tdvovlosggur1fg5@4ax.com...> Eric Jacobsen <eric.jacobsen@ieee.org> wrote: > > >As has been stated, the vast majority of analog FM receivers for > >broadcast use PLLs, mostly because it's a tiny circuit with just a few > >transistors compared to other (especially digital) implementations and > >it does work quite well. > > They are not used for demodulation, but as 1st LO. The normal demod > for analogue FM apps is a discriminator.Not at all, just for cheap radios they use discriminators. A good car radio will use a PLL and besides, there are no definitive results comparing I/Q methods with a PLL published to date so we don't know if say IQ methods are worse (they won't be better). F. -- Posted via a free Usenet account from http://www.teranews.com
Reply by ●January 19, 20072007-01-19
"Fitlike Min" <Fitlike@naeoption.com> wrote:>Not at all, just for cheap radios they use discriminators. A good car radio >will use a PLLI have never seen one with a PLL demodulator. Even most high-end police radios for mission critical two way communication use the good old discriminator setup, of course hidden inside some receiver IC.
Reply by ●January 19, 20072007-01-19
On Thu, 18 Jan 2007 14:27:19 -0500, Randy Yates <yates@ieee.org> wrote:>Eric Jacobsen <eric.jacobsen@ieee.org> writes: > >> On Wed, 17 Jan 2007 23:00:39 -0500, Randy Yates <yates@ieee.org> >> wrote: >> >>>sampson164@gmail.com writes: >>> >>>> Randy Yates wrote: >>>>> Gentle and Wise Readers of comp.dsp, >>>>> >>>>> Two ways to demodulate FM are: >>>>> >>>>> 1. Find the instantaneous phase (e.g., by examining the phase of the >>>>> analytic signal) and compute the phase differences, i.e, f = dtheta/dt. >>>>> >>>>> 2. Lock a PLL to the FM signal and use the VCO control voltage as the >>>>> demodulated signal. >>>>> >>>>> It seems to me that method 1 is much simpler. Why would anyone do FM >>>>> demodulation via method 2? >>>>> >>>>> Let me state up front that I have an idea, but I'd rather get others' >>>>> ideas first without biasing them. >>>>> -- >>>>> % Randy Yates % "She tells me that she likes me very much, >>>>> %% Fuquay-Varina, NC % but when I try to touch, she makes it >>>>> %%% 919-577-9882 % all too clear." >>>>> %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO >>>>> http://home.earthlink.net/~yatescr >>>> >>>> The SNR threshold is lower with a PLL. Search for "Threshold Extension" >>>> for details. >>> >>>That's it, John - that's what I was looking for. Apparently it's >>>fairly well-known - I found it in [couch] after my original post It's >>>apparently also known as FMFB (FM with feedback), which is referenced >>>in both [couch] and [schwartcommtecniques]. >>> >>>Does anyone know how prevalent the PLL technique is in current >>>demods (TVs, FM broadcast, etc.)? Is it pretty much standard >>>fare now or are they somewhat unique? >> >> As has been stated, the vast majority of analog FM receivers for >> broadcast use PLLs, mostly because it's a tiny circuit with just a few >> transistors compared to other (especially digital) implementations and >> it does work quite well. >> >> So the cost/performance tradeoff is hard to beat, and the power >> consumption is tiny. > >Sure - that makes sense. I guess I was looking at it from a >S/W POV. > >Also, if this helps improve performance for analog FM, wouldn't it do >the same for m-ary FSKs and variants? Would GMSK be improved with this >type of receiver?I'd think so, and, depending on how one looks at a coherent receiver, isn't that generally what's done? i.e., phase lock to the carrier and use the detector output to demodulate? I think for the digital signalling cases the demodulation is often done digitally rather than analog for performance improvement, especially wrt the matched filtering. There seem to be a lot of historical examples of FSK/MSK being modulated/demodulated with analog circuits to carry "digital" data, though. Eric Jacobsen Minister of Algorithms, Intel Corp. My opinions may not be Intel's opinions. http://www.ericjacobsen.org
Reply by ●January 19, 20072007-01-19
Randy Yates wrote:> Two ways to demodulate FM are: > > 1. Find the instantaneous phase (e.g., by examining the phase of the > analytic signal) and compute the phase differences, i.e, f = dtheta/dt. > > 2. Lock a PLL to the FM signal and use the VCO control voltage as the > demodulated signal. > > It seems to me that method 1 is much simpler. Why would anyone do FM > demodulation via method 2?Method #2 makes use of the fact that the bandwidth of the modulating signal is less then the bandwidth of the FM signal. I.e. with the discriminator#1 the noise bandwidth equals to the FM bandwidth, versus with the PLL#2 the noise bandwidth is the bandwidth of the PLL. Due to that fact, PLL provides some better performance. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
Reply by ●January 19, 20072007-01-19
Randy Yates wrote:> Do you (or anyone else) know of other techniques that can push the > threshold extension even further down than a PLL?Yes, of course. Pretty much any good algorithm of the frequency estimation can be used. It will make the FM demodulator which performs better then the PLL. Also, the statistics of the audio signal can be exploited... However the radical solution will be encoding the audio into MP3 and transmitting it digitally. No analog solution can beat that. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
