Cursed circuits #6: reverse avalanche oscillator | Hacker News Reader

Cursed circuits #6: reverse avalanche oscillator

lcamtuf.substack.com

89 pointssurprisetalk27 comments

You can even use it to make a simple audio synth: https://www.lookmumnocomputer.com/simplest-oscillator

The man is a treasure

This is amazing. Now I'm even more upset he scored so low on Eurovision.

His music isn't for everyone, it was sort of inevitable. I love it though, I've got most of his discography downloaded haha

In the real world, a lower score beats a high one, except that one time.

This isn't particularly cursed. I mean we've had avalanche pulse generators for well over half a century easily. It's just exploiting semiconductor characteristics they don't teach you in the usual garbage undergrad textbooks.

> the usual garbage undergrad textbooks

Out of interest, please could you give some examples of textbooks you consider garbage, and some you consider not to be (undergrad or otherwise)?

Garbage -> Sedra and Smith particularly - hate it, anything which kicks you in the nuts right up front with Laplace and networks abstractions, anything from Pearson - have never seen a good one.

Good -> Razavi (Fundamentals of Microelectronics), Art of Electronics, most Jim Williams stuff (AN's and articles), Bowick RF Circuit design. They're actually useful.

I find Sedra/Smith a terrible introduction, but a good reference. It's nice once you've already built an intuition for how things work to be able to go back & build up the mathematical models, but trying to understand the behavior of circuits from the math first is a bad order.

I don't know, I found Sedra/Smith to be an excellent introduction.

But personally I think the whole Sedra/Smith vs Razavi argument is basically vi vs emacs for EEs.

Agreed on Sedra and Smith vs Razavi. Razavi uploaded lectures to YouTube that helped me a ton at the time.

What do you think about the author's upcoming book ("The Secret Life of Circuits")[0], for someone who is just starting out with electronics?

[0]: https://nostarch.com/secret-life-of-circuits

Just had a quick scan through it. ToC is a bit rambling. Content sample isn't enough to judge the whole book. I'll reserve judgement. It looks considerably better than Practical Electronics for Inventors and several orders of magnitude better than Mimms' and Make garbage.

Probably a good start. We shall see.

I was going to write my own book on the matter but quite frankly I don't have the time or energy for it.

> Art of Electronics

I would never recommend AoE to pretty much anybody nowadays. The problem is that AoE happily presents circuits you should almost never use right next to circuits you should almost always use with no indications distinguishing the two.

Well, if it's not an advertised spec on the datasheet, and constrained in the direction you're using it, then all guarantees are off and it's on you to qualify the parts on a batch-by-batch basis.

I think that counts as "cursed" from a design-for-manufacturing perspective.

"After a production run of 12,000 units the TR-808 was no more." https://secretlifeofsynthesizers.com/the-strange-heart-of-th...

Behringer sells a clone of the TR-808 called RD-8. They must have solved that problem somehow. They went as far as cloning some vintage Roland silicon that cannot be sourced, so the transistor issue seems minor by comparison.

> "It’s hard to build an oscillator"

I would beg to differ. It's hard to not build an oscillator. An awful lot of digital design is avoiding building oscillators all over the place.

Incidentally, the coolest anti-oscillator I've seen used the back EMF from a relay to push a transistor past its breakdown voltage, and that was the oscillating element.

I would never have thought to try this circuit for fear of letting out the transistor's magic smoke.

But that's the fun part!

https://en.wikipedia.org/wiki/Pearson%E2%80%93Anson_effect

This is it. Exhibiting of negative resistance.

Delightfully cursed.

I wonder how consistent the breakdown voltages are between manufacturers? I mean, I am sure there is some spec, but is it not just a minimum in this case?

It's one of those characteristics you can't rely on unless you select / bin parts. Also depends on temperature. Which is why no one does this in a production design - it's expensive and unreliable.

While the breakdown voltage of the base-collector junction varies wildly even between transistors of the same type, the breakdown voltage of the base-emitter junction varies very little between planar-epitaxial transistors, because in all such transistors the emitter has the maximum possible doping, which is limited by the solubility in silicon, and the base must also be much more strongly doped than the collector, but much less doped than the emitter, which limits the range of possible dopings. Because in this case the breakdown voltage of the base-emitter junction is determined mainly by the doping of the base, there is little variance between transistor types and between manufacturers.

Typical breakdown voltages for base-emitter junctions are e.g. around 10 volt, while the manufacturers specify breakdown voltages like 7 volt, to have a safety margin.

What would happen if you added a capacitor (to ground) on the base? Could you adjust the frequency?

No. That capacitor would charge at a few volts, i.e. the sum between the voltage on a conducting LED and the voltage on a conducting silicon diode (the base-collector junction).

So the capacitor would initially divert a part of the discharge current from the LED, but later it would discharge itself through the LED. So it might make the current impulse through the LED smaller and wider. Such a capacitor would increase the probability that the transistor would be damaged by the periodic breakdowns, by extracting a big current pulse through the base. Depending on the transistor structure, a capacitor on the base could also stop the oscillations, because the current extracted through the base could eliminate the negative resistance that appears on the voltage-current characteristic in the breakdown region when the base is not connected or it is connected somewhere only through a big resistor.

The frequency is very easily adjustable without changing the schematics, by changing the value of the resistor that charges the capacitor, before the breakdown voltage is reached, or by changing the value of the capacitor.

The RC product determines the charging time, which constitutes most of the period of the pulses.

Thank you!

Reply on news.ycombinator.com