The most notable feature of this receiver is that is almost totally immune to SWBCI, even when run on an AC supply. You might hear just a little broadcast pick up way down in the background. Poor AM rejection is a common problem with DC receivers which use an active mixer and can be quite annoying. The use of an analog switch mulitplexer as the mixer completely eliminates this problem.
The input to the receiver is through Q2, 2N7000 MOSFET switch. This connects to the transmitters Low Pass Filter, on the PA side. R10, a 51 ohm resistor, terminates the Tx LPF, so it has its desired reponce. Q4 is used to short the input to the receiver when transmitting. Q3 acts as an inverter, as the bias to Q2 and Q4 needs to be complementry. The advantage to this circuit over the commonly used series L/C with limiting diodes circuit used for QSK is it has low loss and is independant of frequency.
Values for the input tuned circuit on the jfet pre-amp aren't shown, as they will be determind by the band you want to use the receiver on. A plug in module could be made for multi-band use. The 10 ohm resistor in the jeft source lead prevents VHF oscillations.
The mixer is a 74HC4053 analog switch. This part is configured internally as three single pole, double throw switches. Two of these switches are used to emulate the topology of a double balanced diode mixer and works essentually in the same manor as a DBDM does. This mixer has a very high dynamic range, equal to the chips supply voltage of 5 volts. However, to achive the high dynamic range, the inputs to the switch should be biased to 1/2 the supply, 2.5 volts. The circuit works well with no bias, which simplifies making the input transformer, as no center tap is needed. Not only overload problems are eliminated, but due to the analog switchs used, rectification problems seen in diode mixers are also eliminated, hence the much reduced suseptablilty to AM broadcast interferance.
The output of the mixer apears across a 1K resistor. This minimizes losses in the mixer, due to the relatively high "on" resistance of the analog switches in this chip. The detected audio product is amplified by a differential input amplifier, with a gain of 100. Then the signal passes through two audio band pass filters, with a peak response at 700 Hz, a Q of 5 and a gain of 10. These filters give good, but not great selectivity to the receiver. There is a fourth op amp in the TL074 package which could also be used. I would suggest implementing a 800 hz low pass filter with it, and placing it between the diff amp following mixer and the first band pass filter. That should help remove some of the annoying highs that leak through. A LM386 provides additional gain and low impedance output drive. Side tone is inserted into pin 3, through a low pass filter. A sqaure wave can be inseted into the LPF, from a say a keyer chip.
Muting for the receiver is done by using the inhibit feature of the '4053, which turns off the gates. A jfet switch is also inserted in the audio path, just before the LM386. This eliminated annoying and very loud clicks when the transmitter was keyed. Note that muting is acitve high. Where as the transmitter keying is active low, as shown on the following schematic. In my final version, I added a keyer chip which provides the side tone and has an active high keying output. This mutes the receiver and an additional 2N7000 is used to pull the transmitter keying line low.
The VFO is feed into a transistor amplifier, Q1, which squares up a sine wave input signal, which should be at least 1 v p-p. No VFO is shown, as there are a number of options for a VFO circuit. The ideal VFO would be provided by a DDS chip. Any of the standard analog VFO circuits can, of course, also be used.
If the VFO is also to be used with a transmitter, the VFO should be made to work on a frequency other than the direct operating frequency. This will eliminate potentual pulling and stability problems when transmitting. There are two ways of doing this. One is the heterodyne approch, where the VFO is mixed with a crystal oscillator to produce the desired operating frequency. The other way is to make the VFO operate at twice the desired frequency and divide it by 2 using a "D" flip-flop. In this case, a square wave is produced and can be feed directly into the 74HC4053 mixer, eliminating Q1 and it associated parts. Also, the VFO should incorperate RIT function, which is imperative in the use of a direct conversion rig.
If you've never used a DC rig before, you'll find it to be a real pain in the butt! The main thing to remember is, you need to "zero beat" (and that means actual, 0 Hz zero beat) a station you want to talk to, then use the RIT to get the side tone to peak in the audio filter. The one good thing is if there is a near-by QRMing station on one side of zero beat, you can often eliminate it by tuning to the other side band. Unless your trying to QSO someone in between two stations, then your in trouble! Therefore, DC rigs are best used when the band isn't very busy, like during the day or late at night. Never during prime time or during a major CW contest! You also need to remember to re-center your RIT control before tuning around. That's why a center detent control is nice for RIT. It's easy to find the center.
72, Steve KD1JV