zlacker

Interesting, but yeah, he fought way too hard with the analog part driving the transducers

I get it, it's a hobby project, but I feel the knowledge of analog electronics is getting lost in a way (and yes the analog people are the worse at explaining things - I think I saw YT videos more informative than whole university courses)

>>raverb+(OP)
Author here. It's all digital (square waves!).

There were definitely well-known circuits for driving transducers, but I had to improvise to make a circuit small and simple enough to build 100 of them.

Building new circuits requires tinkering and many tries, regardless of the builder's knowledge.

>>belzeb+11
> well-known circuits for driving transducers, but I had to improvise to make a circuit small and simple enough to build 100 of them

Yes, maybe an Opamp that works at 40V could have worked (with an appropriate feedback config), but your end result looks ok. I'm just saying that putting the transducer in the emitter was a non-starter (as you correctly pointed out)

> Building new circuits requires tinkering and many tries, regardless of the builder's knowledge.

Absolutely, I'm not denying it :) It's a great job nonetheless

>>raverb+K1
> I'm just saying that putting the transducer in the emitter was a non-starter

I don't get it. What do you mean by that?

>>raverb+K1
You need to be more verbose as you say analog tech is pretty unusual knowledge so even though I've fought through the 101 course you still make no sense to me. It's really sad and I think it's a provlem I share with many.

>>belzeb+W4
There are two reasons why you don't put it there, one of which you explained "The NPN capacitor here only cuts out the alimentation of the capacitor, but there's nothing to discharge it.". Yes, the transducer has (theoretical) infinite resistive load.

But another reason (which would be valid for a resistive load as well) is that anything plugged into the emitter will have its impedance "reflected" to the base.

Even if you're talking about a digital circuit you need to think about transistor biasing. A pure capacitive load on the emitter means no biasing.

One of the ways you could work around this is to have an inductor in parallel with the transducer so it would resonate at the frequency you want https://en.wikipedia.org/wiki/LC_circuit (though not very practical to your case) - but ideally this would go on the collector, not on the emitter

Your 3 transistor solution is ok for the most part. Maybe it's doable with only one transistor but probably would need an inductor and/or wouldn't be too efficient.

(Not the best explanation, but this would require some trial and error and actually thinking a bit about the analog aspects of the circuit)

>>raverb+z6
That's smart. Didn't think about it!

Though, the emitter is pulsed and I also rely on the fact that it stops resonating after I cut out the power. Unless my LC oscillator has a low Q, it will keep oscillating after I cut out the supply, which is something I'd prefer to avoid.

>>belzeb+29
> Unless my LC oscillator has a low Q, it will keep oscillating

Correct! You can add a resistor in parallel to reduce Q (but it will naturally "not be perfect" since the transducer is putting energy out as sound waves and it's not really a capacitor.