Component choice for class AB amplifier.

Thread Starter


Joined Jul 26, 2020
Hold down the ctrl key and left click on the graph title ["I(Speaker)"]. It will show you the current.
Ahhh, I spent ages trying to find that. Thank you!

So, with a 0.1uF capacitor at C1, I got nearly 6 hours out of the battery! That's amazing! Far better than any of the IC amplifiers I tried. Compared to the 1 hour and 40 minutes I got from my original circuit at the start of this thread that is a huge saving.

It's not as loud as what I had originally, and it sounds "tinny" but looking at the waveform and current consumption suggests that I can modify the value of C1 to find a good compromise between sound quality and battery life.

Edit - Reading back through the thread this is exactly what @Ylli said in the first post, I just didn't understand it at the time. Also, I misread the comment regarding polarised caps. Re-reading the post I realised I could use non-polarised caps and I have a bunch of these on the bench. To get close to my original sound, I need to use a 1uF capacitor. According to the Spice simulation, with a 1uF capacitor, on average, the current at the speaker is still nearly half of what it was originally.

So this was a great success. Not only have I achieved a solution to my original question but I also learnt (a little) about how to do Spice simulations along the way. Thank you so much for your help.

Now, it works, but I'd really like to understand how/why it works... How does the capacitor and diode form a differentiator and how does it clip the signal? I feel like I really need to go back to basics on this one and look at how electrons move around a circuit.

Thread Starter


Joined Jul 26, 2020
Sorry folks, another question if you don't mind. Am I right in thinking that with the differentitator circuit I still need a snubber diode across the buzzer for back EMF (D2 in schematic below)?
Annotation 2020-07-28 105042.png


Joined Feb 24, 2006
I don't think so because your pulse will have a fast rise time, but a relatively slow decay. The relatively slow turnoff will prevent the occurrence of an inductive kick.


Joined Jan 23, 2018
What is so complicated about using an LM386? And there is a connection that provides a 100x gain option.
But it will also work to use a capacitor as a differentiator, which a differentiated square wave is a spike wave. That is calculus 1, week 2. That will produce your noise with far less current draw.