I'd like to build a simple AM radio (so simple it will only receive a single AM channel).

What is the simplest way to build a single channel AM radio that will output to an 8ohm speaker? (e.g. 1030 AM)

My current thinking is along the following somewhat vague lines, but I'm hoping to make them more concrete.

1. use a piece of wire as an antenna
2. create an LC circuit to tune input from antenna to the desired frequency
3. amplify the signal
4. connect amplified signal to speaker

Aside from the antenna, I would prefer to use "discrete" off the shelf components such as capacitors and inductors of particular values that I could plug into my breadboard.

• It also helps to pick a strong station. I noticed mentioned 1.03 MHz. If you're near Boston MA, that's a good choice since it's 50 kW clear channel. Back in grade school, I could pick it up about 20 miles northeast with just a crystal radio and headphones. I did have about a 50 foot antenna from my window to a shed in the back yard. That helped a lot. Feb 13, 2014 at 22:44
• What's with the close votes? This is certainly on topic, and the OP has shown he has at least given it some thought. It's clear enough what he's asking, and it's well enough written. What's the problem? Feb 13, 2014 at 22:54
• @OlinLathrop I don't like the requirements. "Simplest" without specifying some minimum performance requirement isn't a great specification. This is one of the few cases where I think more subjective requirements are better: I'd feel better if it asked for a "good design for educational purposes", but even better would be if it asked about a specific part of the design issue. It's bordering on a "discuss with me" type question due to the lack of a clear problem, and inclusion of half a solution. Feb 14, 2014 at 2:09
• @PhilFrost - simplest, yet still capable of receiving a single AM channel. That's fairly clear requirements. Aug 29, 2014 at 14:58
• It is ridiculously easy to build an AM radio. Several years ago I built one from a LMC660 quad opamp. I actually designed it to be a guitar preamp, but when I touched the input lead with by bare finger, it picked up the local AM station very well! Oct 2, 2014 at 7:57

Your basic strategy to use a piece of wire for the antenna, into a L-C resonant circuit tuned to the frequency of the desired station, amplify that, then connect to a speaker is on the right track but misses at least one important step. That is to detect the AM signal, or to demodulate the amplitude modulated radio signal to get the original audio signal.

After the demodulator, you should also add a stage to drive the 8 Ω speaker with. That will be a stage meant to produce power, not necessarily amplify the voltage. To skip this, you can make just a line level audio signal and feed it into some ordinary audio amplifier, or powered speaker like you would hook up to a PC sounds card. The powered speaker is really just a power amplifier and speaker in a single box.

The simplest AM detector is just a diode. The output of the diode effectively rides the peaks of the RF signal, and goes up and down with its amplitude. You set up the impedance driven by the diode such that the output decays slowly enough so that it doesn't change much in the gap between carrier cycles, but does change fast enough to follow the modulated signal. Fortunately this isn't too hard since there is a large frequency ratio from audio to carrier. If the highest audio frequency is 5 kHz and the carrier is 1 MHz, then you have 200:1 difference.

A even better way to make a detector is to use a bipolar transistor as emitter follower. You can think of this as a powered diode in this application:

This audio out signal can probably be fed directly into a PC "powered speaker".

• That "detector" won't work unless the carrier level extends into the base-to-emitter diode's knee, where it can at least start to rectify. Here are three examples showing the three basic ways to hook a tank up to a 1070 kHz carrier modulated 100% by 1kHz: dropbox.com/s/vxllec30tfspij6/RF%20detect%20NOT.asc?dl=0 notice that the integrator's output is just an attenuated version of the modulated carrier and that no detection has taken place. Aug 30, 2014 at 21:28
• @EMFi: I was trying to show the concept, not the details of biasing the circuit. I'd bias this circuit by holding the base about one diode drop above ground at fairly high impedance, then AC couple the carrier to the base. Aug 31, 2014 at 11:56

Simplest possible AM receiver that outputs to an 8 ohm speaker:

simulate this circuit – Schematic created using CircuitLab

• extremely poor selectivity
• extremely poor sensitivity
• extremely poor volume
• extremely poor fidelity

It is however, without a doubt, the simplest radio that meets your requirements. It works because the voice coil can pick up RF currents just like an antenna, and these currents will move the diaphragm. It will help to use a very cheap speaker, since these have worse shielding.

You will need to be quite close to the transmitter for it to work however, due to the poor sensitivity. And also far from other transmitters, because it pretty much has no selectivity. Being entirely passive and also being very insensitive, the speaker won't be very loud, and since speakers aren't really intended to be operated in this manner, I'm sure the audio fidelity will be horrible. It is, however, obviously maximally simple.

While not an especially good receiver, it is an instructive example, and it does work. It demonstrates that radios aren't magic, and electromagnetic radiation can interact with just about anything. The only thing you need to receive AM is a detector: anything that responds to current differently in one direction or another.

• Anyone care to explain how this is not the simplest receiver? Does it not work? Or is there something simpler? Feb 13, 2014 at 22:20
• This is missing the detector. Other than something accidentally acting as a diode, this won't pick up AM. Feb 13, 2014 at 22:22
• @OlinLathrop The detector is the nonlinear magnetic materials in the speaker. Feb 13, 2014 at 22:24
• @OlinLathrop I don't argue that it's good in any way, but I think the understanding of how this works (to the extent that it does) is a good step towards understanding how more useful radios work. Arguably, this is the radio, and everything else is just an improvement to address one of the "extremely poor" qualities enumerated in the answer. I probably learned more about radios by accidentally making them with a \$20 Radio Shack kit while my father was repairing AM broadcast stations than I did reading schematics of intentional, "useful" radios. Feb 14, 2014 at 2:05
• A fellow I used to work with lived in Berlin when he was a child. He could use a speaker and a diode to listen to the Voice of America transmitter. Connect one end of the diode to the speaker, hold the other end of the diode in one hand and touch the other hand to the other speaker terminal. Presto! Instant AM Receiver. This of course only worked because he lived near the transmitter, and it transmitted with some god awful amount of power (100KW or more.)
– JRE
Oct 2, 2014 at 6:55

Some of the simplest AM radio receivers were based on the ZN414 - a radio on a single chip in a TO-92 package. (replaced by the MK484).

It is poorly suited to drive an 8$\Omega$ speaker directly (it won't be very loud) but you could add an LM386 amplifier as shown here.

If you really want to work with discrete transistors rather than ICs hiding in transistor packages, you could just follow up that tuned loopstick-capacitor circuit with a diode detector (a germanium such as a 1N34A or Schottky diode such as a BAT54 would be best) and an audio amplifier. The missing part of your list is the detector, which converts the RF envelope into audio.

This is the speaker variant of my crystal radio.

Output impedance of the order of a few hundred ohms necessitates use of a matching transformer for the speaker.

A power transformer, with its inherent thick conductors, is an ideal substitute for a low-loss output transformer.

Trials with transformers rated 230V/24V-2A, 230V/24V-5A and 230V/24V-15A proved that the last one is the best.

Reception of the lone local 200kW 612kHz AM broadcast station, located 20km away, is quite good.

The wire antenna used is just 60' long.

• And for some more details, there's a resonable Wikipedia article on the crystal radio, although no loudspeaker variants are mentioned there, it seems. (In fact the page currently says it's impossible to build a crystal radio powering loudspeakers.)
– Fizz
Jan 1, 2015 at 21:15
• And this is one case where Wikipedia is (obviously) wrong. I even found a transcription (but with original photographs/diagrams) of a 1960s crystal radio with a loudspeaker taken from the August 1960 issue of Popular Electronics. It uses the same transformer idea as the above diagram, as you'd expect, but it's a little bit more sophisticated: has tuning capability, etc.
– Fizz
Jan 1, 2015 at 21:28
• @Nandu I didn't realize AM stations could transmit at 200 kW anywhere in the world. In the US, they are limited to 50 kW. May 1, 2015 at 10:17
• Hi Crosley, Here's a list of radio stations in India including the high power AM ones! allindiaradio.gov.in/wpresources/airstationslist.pdf
– user52238
May 2, 2015 at 4:15

The circuit of my simple AM demodulator :