500kHz Oscillator

Mostlikely won't run on a AAA. Check the VI characteristics and you can see there is only a very small window in supply voltage with a chance to start oscillating.

Well, it works in the sim. I noticed it too, but it seems to work anyway at 1.5V quite surprisingly. You could add a potentiometer to the input to modify the input voltage if you're having trouble getting it to oscillate.

This might be new to you, but a 1.5V voltage source in simulation, is not a AAA battery in reality. Not talking about the BJTs.... So build it, test it and see what it does. It was you who said "can run on a AAA battery"!

I know that AAA batteries can't be assumed to be an exact 1.5V voltage source, when they're first bought they're just above 1.5V and as they're used, they drop, and when power is being drawn from them, they're voltage drops because of internal resistance, but at these power levels I'd imagine that they'll be pretty stable in voltage. You could also use a "used" battery, because actually the lower the voltage the more it wants to oscillate, except when it gets too low. It doesn't seem to oscillate around 1V or lower. Don't quote me on that exact figure. Also, no chance I'll build it in real life, as a 12 year old, the closest thing to making my own circuits I have, or will get anytime soon, is my electronic playground, where, unfortunately, I can't get any of the radio circuits to work, but, that's another topic.

I changed the description/circuit so that it specifies "used" AAA batteries.

What hurz means is that that tunnel diode model has the "strange part" of the curve at nearly 1.5V so anything lower then that voltage can't make it oscillate

With a better BJT model, closer to reality e.g. BC547B, its much lower and not even close to 1.5V its below 1.3V NiMH ... whatever.

I see. It must be the saturation current, when I increase it on all parts equally, the negative resistance zone moves to a lower voltage spot.

No, the point is, you have to TEST things before you recommand to use them. Keep it up!

Well, as mentioned above, I have a slight problem with the testing-them-in-real-life bit.