Our notional RADIG Configuration is nothing more than two direct conversion ADE-1 Double Balanced Mixers that on receive convert the received signals to I & Q signal channels. (I = In Phase and Q = Quadrature, out of phase by 90 degrees ). In the early days of SSB Transmitters the phasing method did this same trick.
Now on transmit the ADE-1 act as mixers to take I and Q audio signals and mix those with a local oscillator to provide a signal output at the operating frequency. Then it is with a simple hybrid combiner transformer that we produce the desired output.
A common Band Pass Filter used both on transmit and receive provides the signal filteing to assure we are transmiting/receiving on our band of choice.
The ADE-1's to operate properly are fed RF signals via our hybrid combiner out of the Band Pass Filter so essentially there are two signal channels. Like wise the Local Oscillator input (at frequency) must be quadrature so the you will have I and Q audio signals coming out of the ADE-1's. The I and Q signals are fed to isolated Modem Transformers so we reduce the possibility of hum. Remember this is nothing more than Direct Conversion Receiver technology and one of the problems is HUM!
To generate a quadrature LO so that the signals are indeed equal and 90 degrees out of phase we have a simple logic IC called the Dual D Flip Flop, specifically the 74AC74. When a signal is applied to this IC the outputs are square waves 90 degrees apart and at 1/4 the input frequency. Therefore the input signal must be 4X the operating frequency.
As an aside I mentioned about the early phasing SSB transmitters. One of the issues was that the phasing networks were all resistors and capacitors. Typically they were odd values and subject to aging (read changing values) thus there was frequent adjustments necessary to maintain the 90 degree shift. The IC replaces all of those resistors and capacitors --and no drift and no adjustments. One of the diagrams below shows the wiring of the 74AC74. You certanly don't know what is in a 74AC74 and cannot measure resistors and capacitors; but you know it works!!!
The photo below that shows a second prototype "detector" board. The empty IC socket is for the 74AC74 and the board has the on-board 5VDC regulator for the 74AC74. The four pads to the left of the ADE-1's is for the Band Pass Filter that was a two section. The new standard is a 3 section so I will just install a board overlay to accommodate the additional stages.
I have chosen 40 Meters for the build. But if you are a clever builder --you could fit this area with a socket system so to change bands simply means plugging a new BPF. The SoftRock transceivers did this trick!
I should mention that the board shown below in the current build (not the second prototype) used capacitive coupling between the ADE-1's and the modem transformers. After some experimentation with other coupling methods, while not the ne plus ultra, that is what is installed and seems to perform quite adequately.
The photo of the Band Pass Filter is a three stage that I "lifted" from N2CQR. Given that we are not using any real hardware elements such as a crystal filter, I believe we need a better BPF. It is imporatnt for you to invest some time learning LT Spice and simulate this filter. R1 is for simulation purposes and not used in the final design.
There is a photo in the mix of a block diagram / wiring schematic of how our pseudo Dual Gate MOSFET amplifier stage is connected as a single pass amplifier that is used as the Receiver RF Amp stage on receive and as the Transmitter Pre-Driver Stage on transmit. I have used this in many of my hardware (mostly) SSB transceivers and it is as slick as anything.