A New Line of Transceivers ~ DifX
Transceiver Architecture 2.02In our last post we hit the highlights of the KWM-4 and in the closing paragraphs I mentioned the band switch decoder and how using the three digit code provided by the K5BCQ digital VFO that automatic band switching was possible. Today we have with the Arduino far more efficient means of doing this; but this was how it was done before the uBitx.
Basically the circuit detects the BCD code and translates it into decimal outputs from 1 to 6 (Six Bands). With each output is a PFET that is "switched on" to provide power to the appropriate Band Pass and Low Pass Filter banks. An Arduino Mega 2560 having lots of pins could provide the 3 digit code or if you wanted to waste 10 digital pins you could do it directly. Also shown is how you could switch the bands using 3 toggle switches.
Another innovation was how the "Push To Talk" was handled including how to key the transmitter for CW. This was a total "in house design" and I think one of the special features that are found in the KWM-4 again pre-dating the uBitx.
Shown below is the schematic for the control board. Noteworthy some of this same circuitry is found in the Big Kahuna. On SSB the PTT trips the 4N35 opto-isolator and that toggles the SN74LS000 so that the trigger signals shift from the receive side to the transmit side. It is all DC switching, so no big clunky relays with back emf. There is a separate solid state switch to key an external linear amplifier. [The 4N35, SN7400, 2N3904's and the TIP32C's form the basic control functionality in the Big Kahuna. If the CW capability were added to the Big Kahuna, then virtually the entire same circuitry shown below would be required.]
CW on the other hand is a more complex process and this involves the use of an NE555 timer --not for timing per se but to supply a voltage for a fixed duration. The 4.7 Ufd cap is part of the timing circuit so that the CW oscillator is held on for a period of time determined principally by the value of this cap. Most high speed CW ops prefer a shorter cycle and so values down to 2.2 or 1 Ufd would be used.
Here is the CW sequence. You tap the key and two actions take place the first of which is to start a timing cycle by placing a "high voltage" on the NE555 pin #3. This closes two relays the first of which is a relay whose contacts are in parallel with the PTT switch and the second is the relay that takes the 1st Bi-lateral amp which is normally connected to the Collins filter for receive and SSB transmit but now is connected in the transmit mode to the output of the buffer amplifier. You will recall that the CW signal on transmit does not go through the Collins filter. Tapping the key also keys the buffer amp and keeps the NE555 in the On state. Let up on the key and let the timing cycle complete and the rig is back in the receive mode. Voltage off of Pin #3 also is the source voltage for the CW oscillator. This is quite a complex switching and control system and distinctly feature rich and perhaps far different than other rigs.
It is a N6QW design! Also keep in mind this was designed many years ago and while the Arduino capabilities today would render some of these design elements as being antiquated --it was and is a successful control system. Certainly not QSK; but that was not the design intent. The main design problem was CW with offset and how to simply tap the key and transmit CW all automatically. As with the Collins KWM-2, the panel mounted Mode Switch had to be set to CW -- this also accomplished setting the CW receive mode to USB. The elegance of the KWM-4 was not just casual happenstance but careful and reasoned thought -- five years ago!
73's Pete, N6QW.