A Working Receiver

I completed a basically working AM receiver using the IF building-block from last week:

MK484 based AM Receiver Lash-up

Note the lash-up with a 3.2768 MHz crystal oscillator, and the a two transistor veroboard AF amplifier I often use while prototyping. The resistance wheel is just acting as an attenuator for the AF signal path.

Some FT23-43 ferrite toroids arrived during the week from Kits and Parts. These tiny things allowed me to make an pretty small mixer without using excessively thin and difficult to manage wire, as I usually would when using ferrite beads at this frequency to get sufficient reactance (i.e. more than two or three turns through the bead means the wire must be very thin).

Small Diode Mixer

The receiver produced a noisy, but quite usable signal from about 10uV of RF at 3.68 MHz:

Detected Signal

Finally I could use the AM transmitter to produce a test signal for the receiver. For this test, the TX was powered from a pair of 6 V gel-cells. The receiver was powered from my bench PSU. My VR-500 supplied the audio signal for the AM modulator, tuned to an FM radio station.

A dedicated AF amplifier was built for the receiver. I actually made it too sensitive, using a design I would normally use for direct conversion receivers. The output of the MK484 is high enough to make such extreme AF gain unnecessary. Removing the emitter bypass on the first stage offered acceptable gain, but a simpler amplifier with only two transistors would be sufficient. I also made the output class-A, with a 2N3904 standing quite a few mA - I may rebuild this module to something less of a hack, more efficient, and better suited to the mV output level of the MK484.

Transmitter talking to the Receiver

Note the small board with an IF can on it, this is a 455-ish kHz BFO oscillator for resolving CW/SSB. This doesn't work as well as a proper product detector, the injection level needs to be adjusted with varying signal strength, but the AGC makes a set level usable over a fair range. Higher injection levels de-sense the IF by activating its AGC. There is little way around this as the AM detector is buried inside the MK484 where you can't access it for BFO injection, and is simply the price you pay for such a simple IF circuit. That said, the BFO works fine, I was able to receive the ARNSW morse beacon on 3.699 MHz using this receiver and the BFO. Some success was obtained from injecting the BFO into the mixer LO port, requiring somewhat less injection level adjustment. I would like to avoid a front-panel BFO level adjustment, but a simple switched pot would be a practical solution if you don't mind the extra control.

Some VHF break-through interference was observed with the IF stage during testing. Vega FM at 95.3 MHz would be heard in the noise, extremely distorted. I assume the fairly long input lead from the MK484 to the IF can pad is picking this up, nothing else in the circuit changes the effect when touched. Shielding will correct this, but if I build this kind of IF circuit again I'll be more careful with the layout of this rather high impedance point.

As currently lashed-up the front-end allows anything into the mixer. This means it can harmonically resolve signals, for example, Radio Australia's monster signal on 6.020 MHz made it impossible to listen for ARNSW's morse beacon 3.699 MHz until after the shortwave station moved to another allocation for the day. While their respective IF frequencies where 100 kHz apart, the receiver selectivity was not sufficient to handle the enormous shortwave signal, some front-end filtering will take care of this. A ceramic filter in the IF path would improve the selectivity, or a Q-booster on the IF resonator.

By tuning around with the IF can, the receiver as-is makes a usable shortwave receiver. The AF amplifier can produce ear-splitting audio into headphones, and pretty room-filling audio using a matching transformer into a small speaker. You just slot in different crystals and tune around with the IF can, but doing so will eventually take its toll on the IF slug and its plastic threads. It would be a simple matter to replace the crystal LO with a VFO and tune that way. In fact, once I improve the selectivity I may build a copy to dedicate to SWL. The very trivial nature of the circuit would make it an excellent project for novices or foundation calls (or whatever we call newer hams now days).


Parent article: "2007 80m Homebrew Challenge".