Simple BJT Amplifier

[sshoptaugh1991]

I imagine this has been answered before, since this is a common beginner project, but I just have been having a tough time figuring out what to do.

I am trying to build a single BJT audio amplifier to drive a speaker.  I am using a common 2N2222A transistor, and the speaker is an automotive speaker, 45W rating and an impedance of 4-8 ohms.  My input is meant to be my iPod.  When I build the circuit, it works as expected, but it is just pretty weak and low volume.  I understand a single transistor isn't necessarily sufficient.  Should I try a Darlington transistor, like a TIP120, or should I make a 2-stage amplifier using two 2N2222A BJTs?  Below is the LTSpice simulation and schematic.   From the output voltage, I would think I should get nice amplification, but there is just such low volume through the speaker.  What should I do?

NOTE I know next to nothing about transistor circuit design.  I took a schematic I found elsewhere, and I started messing with values.  The only thing I understand from a tutorial is that the RC network of R5 and C3 is supposed to affect the gain or something.

[moffy]

Use a purpose made IC like the like the TDA7297. I bought a kit for $4 off Ebay. Uses a single supply, say 12v. Distortion figures are not fantastic but they are quite reasonable.

[sshoptaugh1991]

I will certainly give that a shot!  I was put off of IC solutions after using a LM386 to try to accomplish this, but I got nothing but severe distortion (I assume since the minimum gain is 20, and the input was out of my iPod, 1.5V at max volume, while powered by a mere 9V battery, I greatly clipped the signal).

I will try that TDA7297.  Thanks!

[Audioguru]

I was put off of IC solutions after using a LM386 to try to accomplish this, but I got nothing but severe distortion (I assume since the minimum gain is 20, and the input was out of my iPod, 1.5V at max volume, while powered by a mere 9V battery, I greatly clipped the signal).

Any amplifier will produce severe clipping distortion if the volume control is set too high. Didn't you use a volume control?

A volume control is supposed to have a logarithmic taper so that it matches the logarithmic sensitivity of our hearing. When a log volume control is set to halfway then the signal is reduced to 1/10th to 1/20th.

[mikewhy]

I imagine this has been answered before, since this is a common beginner project, but I just have been having a tough time figuring out what to do.

I am trying to build a single BJT audio amplifier to drive a speaker.  I am using a common 2N2222A transistor, and the speaker is an automotive speaker, 45W rating and an impedance of 4-8 ohms.  My input is meant to be my iPod.  When I build the circuit, it works as expected, but it is just pretty weak and low volume.  I understand a single transistor isn't necessarily sufficient.  Should I try a Darlington transistor, like a TIP120, or should I make a 2-stage amplifier using two 2N2222A BJTs?  Below is the LTSpice simulation and schematic.   From the output voltage, I would think I should get nice amplification, but there is just such low volume through the speaker.  What should I do?

NOTE I know next to nothing about transistor circuit design.  I took a schematic I found elsewhere, and I started messing with values.  The only thing I understand from a tutorial is that the RC network of R5 and C3 is supposed to affect the gain or something.

A couple of notes, and then I expect you to find a good book, read up on single transistor amplifiers, and then come back and explain how you improved it.


The common emitter gain is set by the ratio of collector and emitter resistors. Tapatalk won't let me see your schematic as I write this. I recall it being near unity. Pick the collector resistor for say 1 mA, 4.5V drop at the collector. Select the emitter resistor for gain of 3x.

Figure the 9V supply is good for about a half watt into 8 ohms. Consider buffering the output with an emitter follower. An 8ohm speaker will draw 250 to 500 mA. Ohm's Law will guide you to the numbers.

Sent from my SM-T810 using Tapatalk

[Cerebus]

Add the 8 ohm speaker load to your simulation and run it again. (Just connect up an 8 ohm resistor wherever you actually connected the speaker.) It ought to become immediately obvious what is going wrong here. (Hint: See what size and where all the currents are by hovering the cursor over the parts. Compare it to the version without the 8 ohm load.) One further thing that will become apparent is that you need a big DC blocking capacitor between your output and your speaker.

That was your big problem (for this kind of amplifier, I'll get onto that in a minute). The list of other problems this circuit has includes:

• Biasing. The collector's quiescent voltage is set way too high (about 7.3V). It needs to be at 4.5V (i.e. 1/2 V_cc) for maximum output voltage swing.
• Biasing. The transistor's quiescent collector current is set too low (about 1.7mA). This is a class A amplifier, all the current for the load comes from the collector resistor, all the transistor can do in a class A amplifier is divert current away from the load. The collector resistor needs to be as low as practicable.
• Transistor choice. The 2N2222 is rated for 600mA absolute maximum* continuous collector current. The manufacturers rate this part for an absolute maximum power dissipation of 1.5W. That figure is actually unrealistically high -  it is in a plastic TO-92 case so there's no way that is achievable in practice as you'd never be able to shift the heat away from it fast enough. You'd be better off choosing a small power transistor such as a BD139. The range of possible choices is massive and I pick the BD139 because it's common and cheap and about the right power rating (with minimal heat-sinking) for what you could actually deliver into an 8 ohm load from a 9V supply (a bit less than 1.25Wrms at the speaker).
• Extraneous components. You don't need R5 and C3. That type of bypassing an emitter resistor is used when you need (for biasing reasons) AC gain to be much higher than DC gain. That isn't going to be the case for this amplifier.
• The input impedance is way too low - as it is currently biased.
• The output impedance is way too high - both as it is currently biased but also intrinsically too high for this application.

The last two points are the nub of this. You have chosen the wrong kind of amplifier for this job. A speaker is a low impedance load, it needs driving from an amplifier with an intrinsically low output impedance such as an emitter follower. By comparison a common emitter amplifier has an intrinsically high output impedance, dominated by the impedance of the collector resistor.

The usual choice for an amplifier to drive a speaker is a class B amplifier, but they are more complex and harder to design. They are chosen because they are more efficient than class A amplifiers (such as the common emitter amplifier and emitter follower) and waste less power as heat.

If you want to stick with this, I'd suggest that your best choice would be to use a single transistor common emitter amplifier with a voltage gain of between 2 and 3 (for your 1.5V peak input), followed by a simple single transistor emitter follower amplifier. The first stage will give you the voltage gain you need (and some current gain), the second stage will give you the current gain you need to drive a low impedance load like a speaker.

If you want to take this further and design your own circuit I suggest you read the first two chapters of the fabled The Art of Electronics (either the 2nd or 3rd editions) by Horowitz and Hill. That'll take you a couple of days or so. Once you've done that, revisit this and much of what I've said will have become  clear to you on your own.

* Absolute maximum, in this context, means a figure that if exceeded with damage the part. It's a limit to be avoided, not a target to work towards achieving. Good, practical, circuits de-rate components to keep them working as far away from maximum limits as is practicable.