# Help with a multistage NPN BJT audio amplifier

I am a beginner in electronics.

I have recently learned about BJT amplifiers.

I have designed this simple BJT amplifier on my own.

The book I am following only has theories about various topics but I don't know where to start to build a practical circuit on a breadboard.

I mean how can I find out the values of resistors I should use?

As far as I have learned the Vce of the transistors should be around half of the supply voltage for a distortion-free amplification.

My calculations so far:

Assuming $$\I_b\$$ to be very small the voltage at the base of $$\Q1\$$ is $$\1.5V\$$. $$\I_e\$$ at $$\Q1\$$ is approximately $$\\frac{1.5V-0.7V}{1k\Omega}=0.8mA\$$.

Assuming $$\I_c \approx I_e\$$ .$$\V_{ce}\$$ is around $$\(9V-0.8\cdot 5k\Omega)-0.8V =4.2V \$$ $$\r^{'}_e=25mV/I_e \approx 31\Omega\$$

The input impedance of $$\Q1\$$ is around $$\31\Omega\$$.

The output impedance of Q2 $$\(\beta=250)\$$ is $$\R1||R7||R8||Z_{in(base)} = 1.2\Omega\$$.

I think that's the reason I am getting such low output.

• Is my calculation upto here correct? - How should I increase the output impedance?
• What parameters should I consider for the second transistor?
• How should I select the values of the capacitors and resistors for better results?

I always mess up if I try to increase, say, the output impedance as I mentioned cause some other problems like not enough voltage at base.

Note: Resistor R4 is 5k. Sorry for the mistake.

I have chosen the value of capacitors without proper calculations.

I am so confused on how should I select the values of the bypass capacitors as well as the filter capacitors $$\C4\$$ $$\C3\$$ and $$\C2\$$.

I built this circuit but it didn't work as expected. It appeared as though some of the part of audio was cut and filled with distortion.

Since I am beginning to learn electronics I want to start with simple circuits and don't want fancy circuits which use $$\LM386\$$ amplifiers.

Since I want to get the most out of this circuit type, what rules of thumb should I use to select the resistors and capacitors value?

Calculations regarding these will be very helpful for me.

• If you want to drive a speaker, it might be a good idea to add a bipolar emitter follower on the output. Speakers take a fair bit of current; it might be easier to drive an earphone of the sort old cat's-whisker radios used. I've not done any in-depth analysis of your circuit, by the way, so it could be that you just need to change a capacitance or two and it'll work fine--your capacitors do look quite large for what you're trying to do. Commented Feb 6, 2021 at 20:24
• And that's the problem. I don't know what capacitors will work well. I don't want to try every value and experiment so much because its better to learn a few rules and process than to just blindly do it. Commented Feb 6, 2021 at 20:31
• Well, I see one problem right off the bat: CE2 is shorting high-frequency to ground at Q2's base, meaning very little of your signal will actually get to Q2. I'm assuming you meant for this to be an emitter capacitor? Commented Feb 6, 2021 at 20:35
• But really, this circuit has much too high an output impedance to drive an 8-ohm speaker. This is why I suggest a bipolar emitter-follower stage. Commented Feb 6, 2021 at 20:37
• CE2 should probably be in parallel to R3 rather than R8.
– JRE
Commented Feb 6, 2021 at 20:47

Your goal is to have an amplifier that can drive a typical speaker given a typical audio source. The most important early steps to take are:

• Learn everything you can about your audio source (or range of audio sources.)
• Work through the details you've learned about your audio source and refine and document the important details that are more meaningful you. Also document those details you place at a lower priority. Then cross-check your own specifications against what you can find as professional specifications for audio sources and include/modify your specifications, as appropriate.
• Work through the details you've learned about your output system and refine and document the important details that are more meaningful you. Also document those details you place at a lower priority. Then cross-check your own specifications against what you can find as professional specifications for supported output devices and include/modify your specifications, as appropriate.
• Select an appropriate first stage topology that handles your input audio source(s) well. The options here will be constrained by the details of your input source(s).
• Select an appropriate final stage topology that handles your output system device(s) well. The options here will be constrained by the details of your supported output device(s).
• Work out how the specifications for the remaining intermediate stages that provide the necessary gain, bandwidth, etc. These stages will connect the 1st stage with the final stage. These will likely allow a wider range of options.
• Make sure that you provide for negative feedback. This is crucial for: (a) lower distortion; and, (b) mitigate the effect of part variations on the DC operating point; and, (c) cope with ambient temperature variation; and, (d) cope with elevated temperatures due to active power output; and, (e) better establish a known voltage gain. Running a system "open-loop" (no negative feedback) is much more likely to result is inconsistent behaviors.

Some amplifier systems accept a standardized input source type. Others may accept inputs from specific types of transducers (externally or internally powered electret microphones, etc.) Each distinct type of source may require some careful thought in order to properly design that 1st stage. Imagine a specific type of microphone. It will have a certain "optimal" mode of operation and the 1st stage that accepts this microphone's signal must do its best to properly handle that microphone. Some electrets (cheap ones, especially) require $$\500\:\mu\text{A}\$$ to about $$\1\:\text{mA}\$$ of supply current in order to operate, for example. If you plan to support those, then you'd need to study them and work out your own means of supplying that current, properly, while at the same time providing mitigating noise and performing a buffering and/or voltage gain. You can't just slap down any old stage and expect it to work well, or work at all. You do have to think about it.

The same thing goes for your speaker and/or headphone output. Speakers usually require a lot more power than a headphone. Do you want to support both? Just one? It matters a great deal. If for no other reason than you can get by with a much lower power circuit if you are only driving headphones and not speakers. And lower power often means "lots easier to do."

Transducers are devices that either convert physical variations (sound pressure waves, for example) into electrical signals (voltage variations or current variations or resistance variations) or else that convert electrical signals back into physical variations of some kind. The most important (and often the most difficult) parts of a design will be those parts that optimize the transition between those physical variations and their electrical signalling. These are the 1st and last stages. All of your mental energy should go there. The rest, the stuff in between, is usually more of a cakewalk.

Your system uses (ignoring some errors in judgment) two identical stage topologies. So long as you are using preconditioned input sources (headphone jack output from a cell phone, for example) to your amplifier system, your choice may work acceptably though you still need to deal with source impedance and 1st stage input impedance questions well. But it's a terrible stage type to use as the final for a speaker. Your output impedance is very high compared to the load and the delivery of power, which is absolutely important when creating pressure waves in the air, will be quite poor. Barely noticed, if at all.

You need to rethink at the very least your output stage. It's not "fixable." (Well, it's beyond any reasonable modification though there are some crazy-minded modifications that could make it deliver more power.)

For some thoughts, you could look here. It's an EESE answer I gave a while back that may be of some use -- at the very least in helping you see what kind of output stage may be appropriate.

• Thanks for giving such a detailed answer. Commented Feb 7, 2021 at 6:50

Your output stage has idle state Ic =0,8mA

The AC amplitude of the non-distorted output can be max. 0,8mA

AC current 0,8mA peak brings to 8 Ohm about 2,6 microwatts power. That's not especially loud although it can be audible in quiet environment with a sensitive speaker.

Calculations: P=0, 5 * R * Ipeak^2 = 0,5 * (8 Ohm) * (0.0008 amps)^2 = 2.56 millionth part of a watt = 2.56 microwatts = 0.00256 mW. That's about 56 dB below 1 watt. If you happen to have a decent sensitivity small speaker which makes noise level = 90 dB over the standard hearing treshold at 1 meter distance you'll get generous 34 dB audio noise level which is only 1 dB less than what's considered to be good enough silence in usual living rooms at daytime. In bedrooms in sleeping time it's 4 dB too loud as a silence.

The design should start from the wanted power and gain.

• How did you get the 2.5mW can u please show the calculations. What power will be decent for that speaker? Commented Feb 6, 2021 at 23:10
• Smallest radios with a speaker have maybe 0,25W output power. I have a speaker amplifier which outputs 500 watts. It's used in orchestra gigs. In 10000 people stadiums that 500W would be not much more than the buzz of a fly.
– user136077
Commented Feb 7, 2021 at 0:01
• With the 100uF capacitor moved to the emitter of the output transistor the maximum power in the speaker is 2.6 micro (thousandths) of a Watt which is nothing. A 3 or 4 transistors push-pull amplifier with a 9V supply can produce maximum undistorted power of 0.8W into an 8 ohm speaker. Commented Feb 7, 2021 at 0:20

Thumbs up for trying to build the circuit out of discrete components, and for trying to understand how these things work. If you'd like to learn how these things are normally done, in principle, a good google query might be class AB output stage. For a start, look for topologies that have an output "totem" of NPN+PNP transistors connected by their emitters to output.

As for reading, others at this site have introduced to me an excellent free book called Designing Analog Chips by Hans Camenzind.

And... apologies for the self-promotion, maybe I'd like to highlight one response of mine in a recent topic here at electronics.stackexchange. Check out the colored picture and its immediate description. (The topic as a whole does not fit your question very well.) The picture shows a decent example of a class AB power amp with complementary output.

Notice how the class AB audio power amps have no resistors (or very small ones) in the output totem. Thus, each transistor is free to source/sink as much current as it can, if need be. It also has to do with the fact that such amps are operated in a tight negative feedback loop, which precisely controls the extent of how far the two transistors in the totem open vs. close (they work in a push-pull fashion, with a small quiescent current in the middle). For a further explanation of the feedback stuff, see also the very basics of op-amp theory - the basic non-inverting topology.

This stuff is certainly quite a bit more advanced, compared to your classic dual-stage feed-forward amplifier.

• okay i will check out the book. Commented Feb 7, 2021 at 9:56
• Maybe focus on pictures first. Skip the silicon etching stuff and start with chapter 3 on page 59. Constant current sources and current mirrors are a key to achieving (almost) rail-to-rail output swing in the output totem.
– frr
Commented Feb 7, 2021 at 18:21

I made the circuit exactly as you drew it in a simulator and as you say, it doesn't do much of anything.

But when I moved the emitter capacitor CE2 to the actual emitter instead of the base, it worked fine to amplify a 1 mV (peak) signal, but even a 10 mV signal would cause Q1 to clip. This indicates that the amplifier is amplifying too much if your goal is to drive a speaker from a standard line-level output (about 1 volt peak).

• Okay but what other changes should i make to get the most out of it? Commented Feb 6, 2021 at 21:06
• Personally, I'd probably just use a single gain stage for this, lower the gain to something more practical, and put a high-current bipolar emitter-follower stage on the output, probably using a TIP41 and a TIP32, or similarly rated transistors. Commented Feb 6, 2021 at 21:13
• Are my calculations for the output impedance of Q2 correct? Commented Feb 6, 2021 at 21:20
• @shahrozeshahab I think so, but I'm not certain. It's been a long time since I worked with BJTs in depth. Commented Feb 6, 2021 at 21:28