Hi Phil, The following article is a good introduction to demodulating SSB in DSP: https://www.arrl.org/files/file/Technology/tis/info/pdf/98qex003.pdf
It has a bit of maths without going too overboard, and gives you the flavour of how it works. But simplifying as much as possible: - based on the sidetone setting eg. at 700Hz, the DSP will produce a sine wave audio output at 700Hz and add it to the output audio of the radio demodulator. - the frequency of local oscillator in the KX3 is set to the frequency of the signal of interest plus 700Hz, and put through a quadrature mixer. This brings the signal down to a baseband at 700Hz. - the I and Q output of the mixer is sampled by an A-D converter, and the maths mentioned in the article linked above is used to remove the opposite sideband image, and you are left wi the desired signal at 700Hz. When mixed with the 700Hz sidetone, you hear a single tone. - if you tune the VFO a bit, it moves the local oscillator, and the demodulate signal is no longer exactly at 700Hz. You will hear the resulting signal "beat" against the 700Hz generated sidetone. - when you transmit, it will be based on the sidetone frequency, eg. 700Hz. If you have zero-beating in the receiver, then you know the output of the transmitter will exactly match the received signal frequency. If we select CW-R, then we put the LO 700Hz below the signal of interest. If we select other features like IF SHFT=8kHz, dual watch, etc. then the maths gets a bit more complicated, but essentially amounts to implementing a second mixer stage in DSP software. 73, Matt VK2RQ > On 18 Nov 2014, at 11:42 am, Phil Hystad <phys...@mac.com> wrote: > > Don, > > Thanks for the comments. This I guess will be enough. I probably wouldn't > do too much with the DSP math anyway although I have written some digital > filters and playing around with FFT using signals that I invent > mathematically and also some captured audio streams. But, all this was using > Mathematica (Wolfram) so a lot of the heavy lifting is already done for me > (e.g. the FFT and various functions for managing the data). > > 73, phil, K7PEH > > >> On Nov 17, 2014, at 3:13 PM, Don Wilhelm <w3...@embarqmail.com> wrote: >> >> Phil, >> >> The sidetone is a constant audio pitch sent through the audio amplifier >> (once you select the pitch you desire). It is not a function of the VFO, it >> is just a constant tone. >> The demodulated audio (from the signal you are listening to) does vary with >> the VFO setting. >> So yes, the selected sidetone pitch tells the DSP not only what spot pitch >> to send, it also is used to figure out how much frequency offset to use on >> the signal received as a result of the VFO setting. >> >> I can't give you the DSP math to produce that, but Lyle certainly could. I >> suspect the demodulated signal tones and the sidetone are simply added - >> that is at least the result in the analog world. I do know the BFO/signal >> relationship is a "mixer" process and if I recall that in DSP math is some >> sort of multiplication process - as you can tell, I am not well steeped in >> DSP techniques, but I do recall a bit of signal processing math from my >> college days even though that was 54 years ago and I did not use that >> theoretical level of it, so much has become 'fuzzy' with time. >> >> Perhaps some study of DSP algorithms would help you out. >> >> 73, >> Don W3FPR >> >>> On 11/17/2014 5:51 PM, Phil Hystad wrote: >>> OK, so side tone generator is a DSP function. But, your answer leaves a >>> lot of details out. >>> >>> So, if I am using the side tone generator to zero beat another CW station, >>> I do two things. >>> One, I adjust the sidetone pitch to something I prefer. Two, for the >>> actual CW station, I >>> adjust the VFO until the pitch of the CW signal closely matches my sidetone >>> pitch. (I know, >>> I could push the spot button too and I usually do that). >>> >>> If this is true, then the sidetone generation itself is a function of the >>> VFO frequency, that is, >>> it is, say, 600 Hz away for example in a BFO kind of way. >>> >>> I was looking for a description of more of the details of how this was >>> implemented and I admit >>> it is to satisfy my curiosity. Math functions don't scare me though, I did >>> my graduate work in >>> Math. I admit to being a bit weaker with typical DSP algorithms -- more of >>> a philistine than a >>> practitioner. >>> >>> 73, phil, K7PEH >>> >>> >>>> On Nov 17, 2014, at 2:16 PM, Don Wilhelm <w3...@embarqmail.com> wrote: >>>> >>>> Phil, >>>> >>>> The CW sidetone is nothing more than a set of math functions inside the >>>> DSP. >>>> With the proper DSP code, you can make up most any kind of sound wanted, >>>> even sidetone mixed with demodulated signals. That is what the DSP board >>>> firmware is all about (plus a lot more). >>>> >>>> 73, >>>> Don W3FPR >>>> >>>>> On 11/17/2014 4:57 PM, Phil Hystad wrote: >>>>> I am asking this question of K3 and KX3 since I am assuming that the >>>>> answer is the same for both rigs. If it is different, maybe someone can >>>>> let me know. >>>>> >>>>> Question: Is the CW sidetone generated as a separate audio oscillator >>>>> circuit or is it generated from a BFO type circuit? This morning, I >>>>> started thinking of this question and I realized that I don’t know too >>>>> much about how CW sidetone is implemented. I am thinking it is a >>>>> separate audio oscillator but I may very well be wrong. >>>>> >>>>> Follow up Question: If the CW sidetone is a separate audio oscillator, >>>>> what is happening (details preferred) when you use the sidetone to zero >>>>> beat a CW signal. >>> ______________________________________________________________ >>> Elecraft mailing list >>> Home: http://mailman.qth.net/mailman/listinfo/elecraft >>> Help: http://mailman.qth.net/mmfaq.htm >>> Post: mailto:Elecraft@mailman.qth.net >>> >>> This list hosted by: http://www.qsl.net >>> Please help support this email list: http://www.qsl.net/donate.html >>> Message delivered to w3...@embarqmail.com > > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:Elecraft@mailman.qth.net > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html > Message delivered to matt.vk...@gmail.com ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
