Half Of The ZL2BMI DSB Transceiver – A Simple 80M Direct Conversion Receiver :-)

A few weeks ago, Jason NT7S mentioned the ZL2BMI DSB transceiver as a rig I might be interested in building. He was right – I had seen it in SPRAT but for some reason hadn’t seriously considered making it into a project. The mention from him somehow got me to take another look at it and, well, it was such a simple design, it didn’t seem as if I had anything to lose by giving it the old college try.

The great thing about building a transceiver is that if the transmitter part doesn’t work, you’ve still got a receiver. That’s what happened to me (or at least, until I figure out my problems with the transmitter.) Everything was working fine up to and including the building of the driver stage. Once I added the BD139-16 final, I started experiencing problems with a constant residual carrier being transmitted when no modulation was present.

However…….once I had finished the first part of the build, which consisted of building the receiver, I took a little time to enjoy listening to the receiver and generally being surprised that such a small collection of parts would allow me to listen to the ragchewing on 75M. Little things like touching a wire to the antenna terminal and hearing atmospheric noise coming out of the speaker always give me a kick.

Here’s the schematic for my (only ever so slightly different) version of the receiver part of the ZL2BMI DSB transceiver:

There are no RF or AF gain controls in this schematic.  The circuit is still on a board, not yet in an enclosure, and in the experimental stage. If I ever get it into a box, I’ll add an RF attenuator pot in the antenna input circuit. This is even easier to build than a receiver with a VFO, as there are fewer toroids to wind. In fact, there is only only one – the single antenna input inductor.  Coverage was about 3911 – 4009KHz, so I didn’t bother winding a rubbering inductor, figuring that 100KHz of coverage was already pretty good for such a circuit.

The frequency drift wasn’t very encouraging. I was expecting a little better from a ceramic resonator VXO, being around 200 – 300Hz/hour upward drift after an initial 15 minute warm-up period.  The free-running VFO in my 40M NE602 CW DC RX had a better stability – on 7MHz! This board wasn’t in an enclosure though, whereas the 40M receiver was.  I wonder if that could have made the difference?

However, the receiver sounded pretty good, and there wasn’t much to it:

Here’s another view:

Looking at the schematic, you’ll see that as well as the 0.1uF coupling capacitor from pin 5 of the NE602 to pin 3 of the LM386, there are 2 x 0.1uF decoupling capacitors – one from pin 5 of the NE602 to ground, and the other from pin 3 of the LM386 to ground. I saw a wonderful looking version of the ZL2BMI transceiver built by a ham in the Czech Republic, but looking at his schematic, saw that he left out these 2 decoupling capacitors. There was a 0.1uF coupling cap from the output of the NE602 to the input of the LM386 and that was it. I thought that perhaps he was onto something, so I also left out these 2 caps. Well, they are quite important. The 3 capacitors between them form a kind of simple diplexer – as far as I can tell from my limited knowledge. With just the coupling cap, I was hearing stations, but was also hearing breakthrough from nearby strong in-band signals. Adding the 0.1uF from the output of the NE602 to ground cut out the breakthrough as well as cutting down on some of the higher-frequencies in the audio. Adding the second 0.1uF cap from the input of the LM386 to ground helped shape the audio a bit more and cut down on some more of the higher frequencies, making the receiver more pleasant to listen to for long periods.

I shouldn’t admit this in public, but my first thought on seeing the 3-capacitor network that connects the two chips was “How can a 0.1uF cap bypass RF to ground, when the same value is also coupling audio to the input of the next circuit? If the 0.1uF coupling cap passes audio to the next stage, why doesn’t the 0.1uF bypass cap short all the audio to ground?”

After a bit of thinking, I realized that the capacitors form RC filters with the circuit impedances which determine which frequencies they pass and which they don’t. Imagine you’re an audio signal coming out of the NE602 and heading towards the coupling cap for the input of the LM386. You are going to see that 0.1uF capacitance as well as the input impedance of the LM386, which is about 10K. The 3dB cut-off frequency of this high-pass filter is given by:

But what about those 0.1uF caps to ground? Well, they also form high-pass filters, and the impedance in this case is the impedance of the connection to ground which, if the capacitor is connected properly to the ground plane, should be very low. Therefore, the cut-off frequency of this simple filter is much higher. Audio frequencies are blocked and subsequently passed on to the LM386 AF amp, while RF is bypassed to ground.

While pondering the really nice-looking ZL2BMI rig that had been built by the the Czech ham, I decided to do some testing of my own.  Maybe he had a reason for leaving out those 2 coupling caps?  I decided to replicate this in my circuit and found very quickly that without at least one of those 2 bypass caps, the circuit experiences breakthrough from strong stations on nearby frequencies.  I made a recording with and without one of the bypass caps, and here are the results:

If I were going to build this as a simple little receiver to listen to the ragchewing and general chat on 75 and 80M, instead of using a ceramic resonator, I would use a varactor-tuned free-running VFO as with my Hi-Per-Mite DC RX so that I could cover a wider portion of the band. I’d also use a double-tuned bandpass filter for the antenna input and include an RF attenuation pot as well as possibly an AF gain pot.

Naturally, these simple receivers have their limitations, and it doesn’t stop me from dreaming about owning a Ten-Tec RX-340, but I get a real kick out of receiving good-sounding signals from a handful of parts.