The design of  discrete audio amplifier for a 16 ohm load is a bit trickier than suggested above. For any volume with efficiency you will need a push-pull circuit like this one (one for each headphone):

Transistor 1 provides voltage gain. The output impedance is about 1k8 ohm which is too high to be useful to drive a 16 ohm load so transistors 2 and 3 provide "efficient" current gain to drive the speaker.

You can understand why most people just use an LM386 or two instead.

You can drive a 16 ohm speaker with a single transistor with a 1.5 volt supply. The quiescent current would be about 47mA (=0.75v/16R). Assuming a current gain of 100 for a 2N2222 then the bias resistor (R1 below) would be 1k7R [=100*(1.5-0.7)/(0.75/16)]. So this circuit would work (with adjustments):

Problems!

First the transistor gain is unknown with any certainty so you will need to check the collector voltage is actually 0.75 volts and adjust the bias resistor (R1) if not. 

Second, the amplifier may oscillate and/or "motorboat". That can be fixed by adding a 100 ohm resistor between the transistor base and the C1 and R1 connection.

Third, don't change the power supply voltage. For example, 3 volts may blowup your transistor!

Regards AlanX 

Ooh, I really like that circuit. I've seen it before, I just didn't know I could use it here because of that three hundred ohm speaker in the diagram there. I, er, don't really know how to adjust for that.

So... how do I check the collector voltage, and what equation governs the value of R1 (please no calculus! I am not a math geek)...?

Hi Starhawk,

I knocked up the design:

It works (surprise!), I increased the bias resistor (R1) a little to reduce distortion and included the anti-parasitic oscillation resistor (R2) which is only required if the speaker sounds like a motorboat or oscillates. I get a voltage gain of about x6.8.

To test the bias resistor, measure the point "Vc" on the circuit with a voltmeter (between ground and Vc). If it is too high say greater than 1 volt then reduce the bias resistor (say 1k), if less than 0.5 volts then increase the bias resistor (say 3k3). You ideally want Vc to be 0.75 volts

Regards Alan

Just to add to my last comment:

If you want volume control add the potentiometer to the input side (~270R). The wiper terminal to the amplifier input (we don't want to short the audio source to ground).

If you want to run 8 ohm speakers use R1 = 1k and R2 = 50R (i.e. halve all the resistor values).

Is this design the answer to simple audio amplifiers? No not really, it's a power hog at higher supply voltages (e.i. idle power is 2.5W at 9v supply - not practical for a 9v battery!).  

Regards AlanX

Way cool :) What are the values of R1 & R2 for a standard pair of headphones? (I don't 'do' earbuds, BTW, just regular headphones.)

Hi Starhawk,

Headphones can vary but my guess would be 16 ohms is the most common.

For this amplifier it will probably not matter too much. The transistor is operating in an area where the gain falls off pretty rapidly as you try to increase the collector current.

At the end of the day these simple circuits need to be turned up. This is the problem with simple designs and why you don't see them in commercial applications.

AlanX

The comment above is spot on.

Disconnect the positive side of your battery, and connect the negative to the lines labelled 0V on both sides.

Then connect the positive of the battery through a 2.2K resistor to the collector of one transistor, also through a second 2.2K to the collector of the other transistor.

Then reverse the capacitor and resistor on the base inputs: The resistor should connect to the base directly and not through a capacitor, the input should go through the capacitor and then to the resistor/base junction.

This will get you most of the way there, but the amplifier will still distort a little when the input is near zero volts.

Honestly, your circuit has so many things miswired that I would recommend you start over from scratch.

Start with this circuit and get it working on one channel, then make another one for the 2nd channel, and connect the grounds and Vcc together for both circuits.

There are lots of simple 2N2222 amplifiers, do an image search on google and choose one that you like.

I like the schematic there. Can you give me suggested values for the three resistors and the three caps? I'm having trouble visualizing the circuit you describe in your text.

I'm not too sure (electronics novice), but that is designed for 1.5 volts... are you using 1.5 volt batteries (AA or AAA)? It may be overvoltage or something, never built an op-amp or anything using discrete transistors. Also, *look at the pin orientation for the transistors*. That could be it... just pitching in my thoughts, I haven't read through the whole post. Correct me if I'm wrong with anything, constructive criticism is helpful.

There are a few issues that jump out here.

1) There is no DC bias path through the collectors - the 1500uF caps block DC, so the transistor passes no standing DC current through the collector-emitter path, so it can't actually amplify.  There generally needs to be a DC path from the power supply to the collector in a common-emitter amplifier.

2) The potentiometer where it is greatly increases your output impedance.  With the amp "turned all the way up", the output impedance is around 10k - far too high to drive even a 16-ohm load.  I'd suggest putting the volume control on the input of your amp instead, where it would deal with higher impedances and lower powers.

3) I looked at the original you based this design on.  It passes the standing class-A current through the speaker/headphones.  This isn't a great thing - it might work for very low powers, but you have to watch out for dissipation in the speaker coil.

The original circuit you based this on reminds me of early transistor amps published in the 1950s - it's simple and cheap, but performance is going to be poor.

These days, if you just want something to work, an integrated amp like the LM386 is a simple, safe bet, although it requires a higher voltage (maybe 6V or so) to work well.  There are other integrated solutions that can work off lower voltages.

On the other hand, it you want the fun or experience of building a discrete transistor amplifier, there are other, more traditional designs you might look into.  Check out Charles Wenzel's designs for a start: http://www.techlib.com/electronics/audioamps.html

I particularly like his audio amplifier designed around a TL431 voltage regulator - that's a hack right there, but it won't teach you much about transistor amplifiers.

The other thing I'm wondering is exactly what are you trying to amplify with this setup?  A line-level audio output?