AM Radio Transmitter

Very Good

Good

Try using a lower output impedance amp for the supply rail modulation. It's not very well filtered right now.

I tried using a BJT, but if the input went too high, and the audio source was strong enough, the 100uF capacitor would discharge through the base and collector to the power supply, presumably damaging the transistor very much. Anyway, the output filter clears any and all harmonics. I guess I could use a higher power MOSFET, but this one does just fine, and I think a lot of the distortion is because the 100pF capacitor isn't big enough, but if it were bigger, it would decrease the frequency response, and might cause beginners to think that it has some role in the LC tank circuit.

Ok, turns out the MOSFET was limiting the output power. I decreased it's resistance a bit.

Try a push pull amp. It has a lower output impedance and it will enable you to drive easily a more capacitive load (such as the 4.7nF cap). This way you can increase the cap to a higher value (470nF) where it can better sink the rf noise. The problem is that the frequency difference between 1MHz and the upper audible range is not so great, and a capacitor filter will either filter out the rf component too little, or will stress some types of audio amps too much at the higher audio frequency range. So make your amp as low impedance as possible where it'll drive the heavy capacitive load easily, and hope for the best :))

Well, even if you remove the capacitor, it doesn't harm the output at all, I just added a small capacitor to prevent the weird things that happen without the capacitor.

A little more efficency with a single diode http://everycircuit.com/circuit/5096653826818048

@hurz, if you do that then you will be sending more current through the base than through the collector. Will this damage the transistor? Also, the way I had it, it didn't matter very much what the carriers voltage was, and the way you have it, if it's voltage increases or decreases there will probably be a big effect. Well, probably.

Ok, I found out that if I replace the resistor with a diode, and decrease the capacitor to 1nF, the same effect is still there, but the extra current through the base is gone. And the strength of the carrier signal probably isn't that important, so I guess your way really is better.

The antiparallel diode makes the circuit act as an RF rectifier (along with the Base Emitter Junction of the transistor). Previously the voltage was being pulled down, below the Vbe activation point of the transistor, by the fact that the B-E junction of the transistor provides no path to ground for the negative portion of the wave. You can substitute the diode with an RF choke with pretty much the same effect in hand. The choke provides a steady path for DC (the rectified current), and a high impedance path for AC, to ground.

Shut up buggzy

Am I correct that a choke is an inductor? If it is, then what value would be optimal?

Yeah a choke is an inductor, and you should choose an optimal value, where it presents a high enough reactance for the AC signal to not shunt it to ground. It just needs to pass any residual DC that may occur after rectification.

And nothing said.... bla bla bla

At 1MHz 1mH should be enough.

I think I like the diode better. But whoever wants to build it can decide for themselves.