Autodyne Receiver

A direct conversion receiver that can also lock to WWV for use as a standard

This is my version of a project that was proposed on the QRP-Tech Yahoo group.  Chuck Adams K7QO proposed an amplifier chain to get WWV's carrier up to a level usable as a frequency standard.  Glen Leinweber VE3DNL dug out an old circuit design called an autodyne receiver.  Chuck & Glen both refined the thing further, and I jumped in with my attempt which is shown here.

First, here's a link to Chuck's schematic:

Here's a pointer to QRP-Tech, which is a moderated Yahoo list:

Here's Glen's famous Elmer 101 site, but he doesn't have his autodyne ciruits up. Check QRP-Tech for that.

The autodyne has a differential pair oscillating on the frequency of interest.  RF from the antenna is applied to the current source connected to the pair's emitter connection.  Mixing occurs, allowing detection of CW as well as other modes.  A "flaw" of this circuit is exploited here.  If you tune too close to a signal, the oscillator locks to its frequency and the beat note is lost.  So if we deliberately lock to WWV, the oscillator can now be used as a frequency standard.

The circuit also makes a fun little 30 meter CW receiver, considering its simplicity.  It's tempting to call it a "three transistor receiver", but you'd need a couple more transistors to get audio to a halfway decent level.  I only have one audio stage in my current version, so output is a little light.  I connect it to a set of powered speakers.  Chuck added an LM386 to get the audio up where it belongs.

PDF of a hand-scrawled circuit


And here it is again:

 I used a CA3046 transistor array in my circuit, the numbers are pin numbers on the IC.  I didn't change much else from what Chuck & Glen had done.  I did change the input transformer that couples the antenna in.  LTSpice showed a low impedance looking into the base of the bottom transistor, so I changed the transformer to suit.  Later I realized that current out of the base from the pair above was giving a misleading calculation of impedance, so the question is open again. The existing circuit seems to work well though, for what it's worth.

Two high gain amplifiers provide audio out and the amplified oscillator signal out for use when the circuit is locked to WWV and acting as a frequency standard.  In an amplifier like this, you generally chose the collector resistor Rc and then make the resistor from collector to base be beta times Rc, where beta is the DC beta.  That biases the collector around Vcc/2.  A nice and very simple high gain amplifier.

In trying to analyze the oscillator, I like to start by seeing where the resonant tank is.  There's only one inductor, 15 turns on a T37-6 core.  Remember that the Vcc (9 volt) line is RF ground, so the top of the coil is grounded.  The effective capacitance in parallel with the coil is 220p in series with 4700p, plus the C6 trimmer's value, plus the value of 22p in series with the capacitance of the 1N4004, which is being used as a voltage variable capacitance diode here.  Calculating the resultant of all those capacitors and the inductance of the coil, you can come up with the resonant frequency which is in the neighborhood of 10 MHz.  Varying the voltage across the diode with pot P1 allows tuning about a 50kHz or so range.  To move from WWV (10.000 MHz) to 30 meters (10.1 MHz) requires tweaking C6 a little. 

P1 could be pretty much whatever you have on hand, subject to not drawing excessive current off the battery.  From 5k to 100k would be alright, but 5k would pull almost 2mA off your battery.  A multi-turn pot works best for ease of tuning.  If you choose a single turn pot you might want to limit the tuning range by making the 22p capacitor smaller.