Transceiver Architecture Part 2.08
The "Math" of the Frequency Display!
Don't you just hate it when you see information without detail and much like the guy at 1600 Pennsylvania Avenue you hear 'Folks You Will Love This". But it still leaves a lingering question --just how did we get here and how do you insure this is not Fake News!
In our last posting we advanced the idea that with dual conversion this presents some unique problems because of the frequencies involved and how to accurately display the true transmitted frequency. Essentially we have about five frequencies we must deal with in our display process --and that is just for SSB. By way of recap we have the incoming frequency, then the 1st mixer which is the tunable Local Oscillator (LO) followed by the fixed frequency 2nd mixer and then the two BFO frequencies that address either USB or LSB.
Mind you this is Pete's scheme -- there are obviously others and better ones. BUT I understand this and I can be assured that I will not be transmitting out of band. In the final analysis there are many roads to San Diego --the objective is to arrive in San Diego.
In my scheme when you shift from USB to LSB the dial will change by 3 KHz without moving the encoder. In going from USB to LSB you will have to move the LO down by 3 KHz to transmit on the same frequency using the opposite sideband. Inconvenient for some -- well not for me. So if this is objectionable to you, then stop here and go write your own code!
There are several actions taking place in the background as you change bands and change from USB to LSB which I will now detail. [Because we are using two conversions and the conversion frequency is above the incoming -- the net effect: There is no sideband inversion. Thus the lower BFO frequency will be used for USB and the higher frequency BFO for LSB.]
- Let us say you will start by tuning in 14.2 MHz on 20 Meter Upper Sideband --a favorite spot of mine. As you put the mode switch into the USB mode and turn the encoder to 14.2 there are some behind the scenes steps taking place. The first is that 1500 Hz will be subtracted from the display formula and secondly the LO is preloaded with a frequency that will show 14.2 MHz on the display. But the display is actually the subtraction of 45 MHz and the subtraction of 1500 Hz from the start up frequency. The 1500 Hz is the nominal offset from the center frequency of the SSB crystal filter. [Typically it is +/- 1500 Hz depending on the sideband. ] For LSB on 14.2 MHz we will be adding 1500 Hz.
- The LO uses a subtractive mix process so the LO - the incoming signal = 45 MHz. Our pre-loaded LO frequency already contained a 1500 Hz add. So our subtractive mix actually resulted in a frequency of 45001500 MHz plus the Voice signal. Since the Bandwidth of the ESC 45 MHz Filter is 7.5 KHz --this is not an issue.
- The second mixer is at a fixed frequency of 56.5 MHz and the second down mix results in an output of 11.498500 MHz + voice.
- Feeding this signal into the Product Detector with the USB BFO of 11.498500 leaves only the audio voice signal which is upper sideband.
- For a LSB signal the preloaded LO will have to be tuned down by 3 KHz to put you on 14.2 MHz LSB. So now the LO is at 59198500 MHz and when we subtract 45000000 we must add in 1500 Hz to make the display read 14.2 MHz. So putting the USB/LSB switch into LSB causes the addition of 1500 Hz to the formula so that the display will read the true transmit frequency.
- Thus the first mix has the LO at 59198500 - 14200000 + Voice = 44998500 + Voice and this mixed with 56.5 MHz 2nd Mixer signal results in a frequency of 11501500 + Voice.
- Sending that signal on to the product detector with a LSB BFO of 11501500 results in LSB Voice.
Thus you have the decode on how to use the dual conversion scheme yet have the frequency display read properly for the true transmitted frequency. CW can be done the same way --but involves a lot more code and since I am not a CW person have not done any more with it. But shifting one of the BFO frequencies for CW transmit will get you there as was done in the KWM-4 and use USB for receive.
Time to start building your DifX. In a couple of days I will post a link to the Arduino Mega 2560 code and that will be on my website at http://www.n6qw.com
73's
Pete N6QW