Guru
Joined: 05/10/2019
Location: United KingdomPosts: 7844
| Posted: 03:03pm 25 Mar 2023 |
Peter's circuit has the same problem that simple ones tend to do - they put the output current through the speaker. What that does is pull (or push) the cone away from its centre position so that it will start to bang against the pole piece (or come out of the pole piece) on peaks in the music. Those peaks will be a lot lower than its normal power handling rating. It's never a good idea to have this sort of thing happening. The answer is to use a power resistor as the load and then couple the speaker across it via a capacitor. Unfortunately that also brings problems. Half the power is lost in the resistor (so it runs hot) and there is less available for the speaker. Also, the value of the capacitor restricts the lowest frequencies that can be delivered.
Some headphones simply won't handle any amount of DC through their coils and distort horribly if there is more than a mA or two. That's because the diaphragm moves such a short distance anyway.
mozzie's PNP/NPN circuit may be a better bet. Something similar, with a biasing arrangement anyway. It might be better to run it from 5V and use a third transistor (common emitter) to drive the output pair. It does start to get a bit bulky though, as do heatsinks for class A. :)