One of the problems frequently mentioned about direct conversion (DC) receivers is their susceptibility to hum. My FOXX-3 let me experience this first-hand--The hum was unbearable when I hooked it to a 12 volt supply. And while it was quiet with a 9 volt battery, output power (both AF on receive and RF on transmit) suffered.

Doug DeMaw (W1FB/SK) in his book "W1FB's QRP Notebook" talks about building a power supply for use with direct conversion (DC) receivers and makes some good points. I partially rebuilt an old RadioShack supply I'd picked up at a hamfest using his advice.

There are really two problems. The first is that RF (from the DC receivers local oscillator) can get into the power supply. The second (an more subtle) problem is what happens to that RF once it's in the supply.

The transformer's secondary typically feeds a full-wave rectifier bridge made up of four diodes. If RF reaches to these diodes, they can act like a mixer. One of the products (an "output") of this "mixer" is the RF (from the DC receiver's oscillator) modulated by the rectified AC. This, in turn, can be rebroadcast and received by the nearby DC receiver.

So to simplify, we've got a big loop. The DC receiver's oscillator generates a signal on the frequency we're listening to. This RF signal gets into the power supply. The rectifier bridge modulates this RF signal--we now have a new, modulated RF signal. This signal is radiated by the power supply (probably through a cord connected to the supply. The modulated signal then enters the DC receiver's front-end. The receiver now "hears" the power supply's hum.

Okay, that either made a lot of sense or totally confused you. Thankfully, the solution is pretty straightforward whether you fully understand the cause or not...

1. I put a 0.01 uF capacitor (rated 1kV) across the AC line as it entered the supply. This keeps RF from entering the supply on the line cord. (The line cord and the house wiring it's connected to can act like an antenna.)

2. I installed two inductors in the direct-current output. I used some from a variety pack I got at RadioShack--probably 10 uH or so. I put one in the positive lead and one in the negative lead where they connected to the output terminals. DeMaw used a ferrite core for this which should work as good. This keeps RF from coming in the direct current side of the powersupply.

3. I placed 0.01 uF capacitors (rated 50 volts) across each of the diodes used in the full wave rectifier on the secondary. If RF gets into the supply in spite of steps 1 and 2, this will keep it from getting modulated.

After these three changes I can actually power the FOXX-3 with the 12 volt supply. I use about three feet of cord to connect them and the FOXX-3 has no noticeable hum. If I get the FOXX-3 within a foot or so of the supply it starts to hum, I'd guess this has to do with induced AC voltage in the transceiver.

When I posted this suggestion on an e-mail list there were several suggestions about what type of cord to use in making the power connection. I use about two feet of red/black zip cord I got from RadioShack to a 2-pin Molex connector (I use the ARRL recommended standard for this) and then about a foot of twisted conductors to the FOXX-3. I use this same configuration for most of my projects, just replacing the last foot with the one appropriate for the end device.

I've never had any trouble with this configuration, but you might want to consider some of the suggestions made...

You can use twisted conductors. The advantage of this is that the twisting will tend to cancel out noise induced in the wires. I almost always do this with unshielded audio cables. For power cables, my personal experience (for what it's worth) has been that twisting the conductors is great for convenience but doesn't really make much difference otherwise.

There was also a suggestion to use shielded cable for the power supply. Again, my personal experience has been that this isn't necessary. If you do use shielded cable for power connections, I don't recommend using coax cable. Cable television companies often do this and find that unless the cable is designed for this type of service the small wires in the braid (the shield) will often get brittle over time.

Things like satellite dishes often carry power to the remote LNA (low-noise amplifier) with few problems. Even RadioShack sells small TV antenna amplifiers that are "line-powered" using a remote injector. So power over coax isn't a sure fire recipe for trouble. But for me... Well, once bitten, twice shy.

If you have supplies in the shack you haven't modified, you may need to turn them off. The house wiring can act like an antenna and carry your RF into other supplies. And the house wiring will also act as an antenna for the "re-modulated" signal from a powersupply.

A lot of people talk about "60 Hz hum". Actually 60 Hz hum itself wouldn't be as much trouble--many cheap headphones and speakers wouldn't reproduce it with much volume. The hum we usually hear is actually 120 Hz (with a lot of harmonics). This is because a full-wave rectifier acts as a primitive frequency doubler. If you don't believe me (it didn't seem intuitive to me when I first heard it), sketch a 60 Hz wave and what a full wave rectifier output of this wave would look like. Count the times the voltage hits the maximum and the number of times it hits the minimum. You'll find they double after full wave rectification.

Sometimes the biggest problem with 60 Hz energy isn't necessarily what you hear, but what it can do without you hearing it. One time I was working on a receiver project and found the audio was badly distorted with a trace of hum. I injected a 1 kHz signal and looked at the signal on an oscilloscope. What I saw was my poor little 1 kHz signal "riding" on a 60 Hz waveform. Filtering in the audio stages kept the hum to a reasonable level on the output, but the huge magnitude at the input overdrove the early audio stages and caused distortion.