Minimal Transceiver Development
In the most recent SPRAT (#141 Winter 2009/2010) Tony, G0EBP, presents a simple two transistor, one audio IC, transceiver. This caught my eye as it was using a MOSFET as the power amplifier (PA) on transmit and a mixer on receive. This use of a component in two different ways intrigued me (something that Steve Wozniak mentions in his book iWoz) and I had also been reading about MOSFET’s as mixers, in particular Steve Maas Microwave Mixers book and Terry, WA0ITP, and Jason, NT7S, Fall 2009 QRP Quarterly article VRX-1 – the soul of a new kit.
The circuit in SPRAT was simple and I had most of the components and so I started to build a 40m version and experiment. Experiments involved changing the oscillator circuit (from a BJT Pierce oscillator to a JFET Colpitts) to get a driving waveform that was more sinusoidal, modifying the muting scheme and attempting to add a little high frequency roll-off in the audio stage (see Sverre’s, LA3ZA notes on the LM386). I have a few crystals for different 40m frequencies and so used a simple header socket so crystals could be changed easily and without soldering.
RF low pass filtering was added and the output waveform was a nice sinusoidal shape (free of harmonics), as can be seen below. I found that I could obtain outputs of about 160mW from 9V and about 250mW for 12V on transmitting into a 50Ω load. This is down on the 1W reported by G0EBP, probably due to the reduced drive from my modified oscillator, but I was happy with the reduced level. G0EBP had a good idea of adding a buzzer for the side tone, which I incorporated.
On receive there was MW broadcast breakthrough if I used my random wire antenna, but if I used the doublet with the Z-match tuner the MW broadcast was filtered out (by the tuner) and became either negligible or none existent. After reducing the broadcast breakthrough the receive audio was not bad. I used the Elecraft XG-2 three band receiver test oscillator to check sensitivity. At 50μV (equivalent to S9) it was easily discernible. A 1μV it was still possible to hear the tone, albeit weak. The main problem on receive is the wide frequency range the receiver has. There is not RIT or offset on transmit, this could be an issue if someone is zero beat with the rig.
I measured the backwave (the oscillator signal leaking out on transmit) and found it to be 100mV into 50Ω, which I calculate to be 25μW (2.5×V² where V is the peak to peak voltage into 50Ω, and the result is in mW, see p7.9 in Experimental Methods in RF Design).
No contacts made yet. Now to box it up and draw the schematic for a future post.