Hartley Oscillator Modification

I don't see a special reason for the second transistor. Just delete it and its working the same. But why not....

It looks funkier this way haha, and it produces two antiphase waves. Someone might need that

Plus, if both waves are symmetrical and you run them through a transformer, you can completely clear out the second harmonic. That may be beneficial.

That would be interesting to see how you remove, by substraction, harmonics and still have the base frequency present.

Not substraction, negation. Like in amplifiers with center tapped transformer all even harmonics are cancelled and all odd harmonics summed.

Substraction, right. But that won't lower the THD at all. And is only nice for the human ear cuz we like more odd harmonics then the even once. At 40MHz who cares that.

How do you determine how harmonics rich a wave is?

By measuring the THD. Or for higher frequencies we use a spectrum analyzer.

And by the way, aren't the even harmonics more pleasant for the ear? I personally like to have a little second harmonic to enrich the sound.

If you take the 2nd and 3rd harmonics of a particular frequency, musically, it produces a perfect 7th. The ubiquitous guitar "power chord" consists of the root(f), the seventh (1.5f), and the octave (2f). Everybody uses them because everyone likes them. As for the circuit, are you certain it's a Hartley? It's certainly hearty, but it hardly seems like a Hartley at heart. Oscillators based on transistor pairs with bases cross-coupled to collectors in this fashion are most closesly related to the astable multivibrator. The Hartley is just the dual of a Colpitts, both of which have a single feedback path as defining features. (as i understand them) I could be wrong. Cool cct in any case.

WTF: You can guestimate the harmonic richness of a wave at a glance by the "sharpness" of the transitions.

Also, and i apologized for spamming you full of comments, but i thought it worth pointing out that inductors *NEED* some series resistance, even if it's very small. Without it, this circuit doesn't require feedback, or even a power supply once started. The math describing the behaviour of the thing will be full of divide-by-zero issues. Keep it real. Peace.