Antenna for broadcast AM reception on ARM Radio SDR

Question

I'm trying to build the ARM Radio project detailed here. I routed and fabricated a PCB for the anti-alias filter and preamplifier shown on page 2 of the linked PDF, and reproduced here:

Unfortunately there's no mention of what antenna I should use, so I'm trying to use an AM loop antenna taken from a boombox, similar to this one:

Unfortunately, reception is awful (trying to tune to the local AM broadcast stations). Mostly I get noise mixed with interference from a local FM station broadcasting from about 1 km away. At times I get the feeling I'm getting some reception but frankly I'm not sure. Injecting a weak (~50 mVpp) AM signal from a signal generator directly into the ARM board's ADC (bypassing the anti-alias filter) works fine, producing a loud, clear and correct output, proving that everything from the ADC onwards (software, etc.) is working. Reception is great with a car head unit, parked in the same spot where the SDR was tested, so it's not a signal strength issue. My best theory at the moment is that I'm not using the correct antenna to match the circuit (as shown above, the circuit includes a transformer and an inductor/capacitor network for filtering and impedance matching).

I'd rather build a new antenna than design and build a new circuit to match my loop antenna. Thus: what kind of antenna should I build to work correctly with the circuit above?

Answer 1

Ok, doing a little work with the calculators found here and here, I came up with an idea that I think you might like.

Materials:

• 14ga-18ga "fray wire"
• 9" section of 4" PVC pipe
• Power drill with ~1" hole saw & 1/4" bits
• 3/4" pvc pipe (at least a foot)
• 3/4" & 4" PVC pipe end-caps
• 1+ zip tie(s)
• Magnet wire (about 150')
• Some BBs, iron powder, etc.
• 50ohm coax cable (to reach the receiver
• A 3' or taller frame to hang it all from

If you ground one end of your primary (50ohm) winding on the input transformer, you can then run an unbalanced coax line to a "loaded quarter wave" antenna that can be made resonant with a 50ohm impedance for our band of interest. FYI: AM broadcast band reaches across nearly 2 full octaves of bandwidth (0.5MHz-1.7MHz), so it's not very well possible to receive the whole band resonantly with a 'fixed' antenna.

Construction:

• Strip the last 1-2" of the coax from the antenna end;
• Solder a 1' length of the 'fray wire' to the coax center conductor;
• Drill a 1" hole in one of the 4" pipe caps;
• Insert the piece of 3/4" pipe through the hole; glue on one 3/4" cap; fill with BBs/iron filings; then slip on the other 3/4" end cap (so the 3/4" pipe filled with metal can slide up & down through the 4" pipe cap, but can't escape).
• Glue the end caps on the piece of 4" PVC pipe, leaving 6" of 'bare' pipe between the caps;
• Drill a pair of 1/4" holes near the piece of 3/4" pipe on the 'top' 4" cap (we'll call the 3/4" pipe section the "tuning slug," as that's its function), to attach the rest of the 'fray wire' to for hanging the inductor (whole PVC pipe assembly becomes a 'tunable inductor' once it's done).
• Solder the free end of the first 1' section of 'fray wire' to a stripped end of your magnet wire, then glue to the side of the bottom 4" PVC end-cap;
• Wrap the magnet wire around the 'bare' 4" PVC pipe (between the end-caps) for about 150 wraps (or 150' of wire, both are 'close enough'), keeping the wrappings evenly spaced (or evenly overlapped) across the 6" length.
• Solder another section of fray wire (about 1' long) to the top end of the (now wound) magnet wire & glue to the side of the top 4" cap; then loop the fray wire through the 2 1/4" holes and tie/glue in place for stability.
• Attach the free end of this top wire to the mounting stand, so that the shield of the connected coax at the bottom can nearly contact the ground.
• Solder a short "ground lead" of bare wire to the coax jacket (when in use, this wire should be grounded to a ground plane as near to the antenna as possible)
• Attach a zip-tie "pretty tight" around your tuning slug, so by sliding the zip tie up/down the length of the slug, you can adjust how much of the slug's metal filling is inside the coil (and thus, control the coil's inductance).

OK, that took a bit more explaining "on paper" than I expected in my head, but if you get this far, you should be ready to "fire her up." Start out with the tuning slug mostly outside of the inductor, then tune to a high-frequency band AM signal in your area & adjust the tuning slug for best reception. More slug in the coil = lower frequency tuning for the antenna, so if you can't get the high-frequency station to tune in 'great' at any slug position, try tuning to a lower station; if that works, remove metal filling from the slug as necessary.

Answer 2

AM broadcast (at around 1 MHz) is easy to amplify, so gain and sensitivity is not your biggest problem. Your observation about noise and another station 1 km away suggest you may have too much signal. Your receiver may be overloaded. Strong signals can drive the front end into nonlinear regions and unwanted signals can mix together to provide all sorts of spurious input.

Before going for the high performance antenna, try a trick or two. Try no antenna, then a short piece of wire. You can also rig up a low-value potentiometer as an attenuator that lets you vary the amount of signal from the antenna that's getting through.

Another possibility is that the amplifier stage is oscillating. That's bad enough but the oscillation will also mix with everything else as described earlier. To test for that I would touch various points in the input circuit and see if it cause big changes. That's not very descriptive but the noise might get a lot worse or it might quit entirely. This is seat-of-the-pants stuff, but I have a lot of experience with this sort of troubleshooting.

Lastly, make sure V2 (the 5V supply) is clean. Taking this from a computer USB port is asking for trouble, at least until the rest of the radio is proven. Use a battery to get a fair test. Once the radio works, you can try the original power source and see if it's clean enough. It might need some RF filtering, which I don't see any of.