1
$\begingroup$

Let's say I'm listening to an AM radio station in the United States at 680 kHz. From my understanding that means there's a carrier wave at 680 kHz and the bandwidth for the station goes from 675 kHz to 685 kHz.

When I move the tuner of my SDR in 1 kHz increments, the AM audio doesn't sound any different until the carrier wave is out of the 10 kHz slice of the spectrum I'm listening to. Once the carrier wave leaves, the audio becomes very garbled, but not high-pitched.

I'm surprised by this, a bit; my naive understanding of tuning slightly off-carrier like this is that, from the radio's perspective, everything changed pitch - the carrier that was at the exact frequency we were tuned to is now at 1 kHz, the spectrum below 676 kHz is gone, and everything is shifted by 1 kHz. Indeed, when I set my SDR to listen directly to the modulated audio, this is what I hear - a 1 kHz tone and everything is pitched up.

I did try to study how AM tuners and demodulators (in general, not specifically for the SDR I'm using) worked, learning about envelope detectors and how they use a simple capacitor and resistor to follow the modulated signal upward and decay relatively slowly. The tuner itself uses low- and high-pass filters (I guess) to clip away anything out of the 10 kHz bandwidth. I also read about carrier recovery, but the Wikipedia article didn't say anything about its use in AM radio.

Here's a video I took of 680 WHBE that demonstrates what I'm talking about.

(sidenote: being off-tune in FM works like I'd expected; the demodulated audio clips as the carrier wave moves in and out of the bandwidth. It's just AM that's not making sense to me.)

$\endgroup$

3 Answers 3

2
$\begingroup$

The super short, super cynical* answer is that if someone invents an AM demodulator that lets the carrier wave through as you are positing, they say "goodness, this is a crappy demodulator" and they don't publish it (unless it's for a class, and they want to get a C instead of something worse).

You can demodulate AM with a carrier tone -- just use a single-sideband radio that's tuned off frequency a bit. You'll be able to listen to the content, and you'll hear the carrier whistling**.

The oldest AM demodulatior is an envelope detector. This detector takes the filtered AM signal and rectifies it with one or more diodes. It works quite well, and because its action is to take whatever's presented to it and output a slow signal that tracks the input signal's average amplitude, the fast carrier wave gets filtered out.

Newer AM radios might use a synchronous detector. This actually multiplies the AM signal with a local version of the carrier wave. If you had a synchronous detector and the local carrier wave were not spot-on the correct frequency, then you'd hear a whistle. But the "synchronous" part of a synchronous detector is that it picks out just the carrier wave from the signal and either locks onto it with a PLL (a "true" synchronous detector) or it filters it tightly and amplifies the heck out of it (really, not a "synchronous" detector but an "exalted carrier" detector -- I'm not going into detail, you can look it up).

* Or super realistic, depending on how you define "realistic" and "cynical".

** If you're listening to music on AM with SSB, unless the tuning is absolutely spot-on it'll sound utterly terrible. In the waning days of the cold war I'd sometimes listen to Radio Moscow, just to see what the Soviets wanted me to think. Listening to the speech was just fine, but the music -- ugh. I never did find out if that music was any good or if it was just my radio.

$\endgroup$
1
  • $\begingroup$ I bet it was the music, seeing that you really can't make an AM channel wide enough for anything remotely (huh, a pun) enjoyable; there's enough multichannel on mediumwave (and shortwave) that DRM+ has some 41 to 107 Hz carrier spacing in its OFDM scheme, and there's nothing with which your receiver could have equalized a full music bandwidth worth of channel. $\endgroup$ Jul 21, 2022 at 20:25
2
$\begingroup$

well, it depends on what type of demodulator your SDR-based receiver uses. But possibly, it's an envelope detector, i.e. what matters is the instantaneous received power, and that doesn't change with frequency offset.

The tuning in that case serves the purpose of putting a filter at the right bandwidth, so to isolate the channel.

(by the way, you say "how envelope detectors use capacitors": You're confusing the principle, envelope detection, with the implementation, an electronic circuit. In an SDR system the demodulation is not done through some discrete circuitry, but simply by applying math to the signal.)

I also read about carrier recovery, but the Wikipedia article didn't say anything about its use in AM radio.

Well, since AM broadcast have an unsuppressed carrier wave, you can lock some algorithm (typically, a PLL), on that carrier and make sure you're nicely centering the signal in spectrum for whatever following processing steps your demodulator needs to do. It would not only allow you to relatively precisely "home in" on the channel (which, as shown above, isn't very necessary if you do envelope detection), but also give you a frequency with which you can shift the signal in spectrum. That way, you can put your channel where your filter is, automatically.

Whether or not your receiver does that: we don't know!

$\endgroup$
1
  • $\begingroup$ I'm sorry, I really needed to clarify the OP. Certainly my SDR uses software demodulation. I was reading about AM radio in general to find out how you'd demodulate it without digital circuits. $\endgroup$ Jul 22, 2022 at 0:49
0
$\begingroup$

I think I figured out the flaw in my mental model. I've been envisioning a standard AM radio as shifting down the incoming signal to baseband - that is, when tuning to 680 kHz, it shifts the range from 675-685 kHz down to 0-10 kHz and demodulates that. Now I see that, especially with envelope detectors, it's not doing any frequency shifting at all. 680 kHz is 680 kHz, regardless of how accurate your tuning is, and as long as the carrier is in range, the envelope detector will be able to decode it correctly.

$\endgroup$
1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.