So i always have been interested in audio amps, and i created this one. It works pretty well considering that the TIP41C is not "ideal" for this aplication, but i want to know how to improve it(get more power out of it maybe), my power supply works up to 30v DC 12A, so not a problem around there, the 4 ohm speaker(R3) is rated for 25W.
Q1 doesn't get hot or anything, Q2 is with a pretty big heatsink, R1 gets burning hot and i am looking to use more resistors in parallel to avoid this.
LED1 is there because without it the simulation doesnt work at all, the original design had a resistor there, but i couldn't figure a good value for it, and it also worked without the resistor... soo i took it out(not ideal i know).
How can i improve this circuit?
The bass is kinda Muddy(RHCP music for testing), but high frecuencies are fine.

 

What is created is a two stage amplifier using a power transistor as the driver, and an LED as part of the output circuit. And certainly R1 will get very hot because a lot of current is passing through it. This is a class A amplifier, and class A operation is seldom chosen for power amplifier stages because it is not at all efficient.
There are hundreds of good amplifier circuits published, and while this amplifier probably works, it would not be among those circuits. To use these transistors in a more efficient amplifier you need to have the two transistors in series with the speaker connected to the middle of the stack, Then, each transistor needs to be biased just barely into conduction. Then the audio signal need to be applied to the bases, BUT one transistor getting the inverted signal while the other gets the non-inverted signal. THAT will require either more transistors or else a transformer, such as was used on the very first transistor guitar amplifiers. So there you have some feedback on how to improve your amplifier.

Unfortunately, when it is improved, it will have only one stage and so you will need to add another amplifier stage.

 

Here's a 20-Watt Amp in a single To-220 package, for ~5.oo.
https://www.digikey.com/en/products/detail/texas-instruments/LM1875T-LF05/3640532
.
.
.

 

How could that LED possibly survive?

 

Do you want to make a functional amplifier, or learnIf you want to get somewhere how to make an amplifier?
If the former, follow @LowQCab 's suggestion.
If you want to make your circuit at least functional, put R1 where the LED is, remove C2 and connect Q2's emitter to 0V.
Then add a resistor between Q1's base and ground, about 2k, to keep it stable as the gain changes with temperature.
Gain is about 15 and distortion is a few percent, but it's an improvement, and R1 will still be hot.

 

The way to improve on your design is to scrap it and go to a complementary push-pull Class AB amplifier.



We were just discussing this recently here.
https://forum.allaboutcircuits.com/threads/complementary-amplifier.184265/

 

The original "amplifier" produced an output power of only 0.0048W and LOTS of heat.

The improved class-A amplifier (and room heater) in post #5 produces an output power of 2W and 34W of heat.
A class-AB amplifier produces an output power of about 5W and heats with about 4W.

A modern class-D amplifier (TPA3116) produces 12W of output power and heats with only 1.2W.
The TPA3116 costs $4.98US at Digikey today but they are sold out.

 

Here you go - this works a treat - 4k7's across the output xtors help reduce crossover effects to very low levels - the 120K res sets the gain - larger for more, smaller for less.p.s. the gain is 21 as given above, so for full output, you need a little less than 500mV rms input.

 

this works a treat

Really?

 

@ Ian0, yes really - feel free to build and test or LT spice it ... if you are able ...

 

@ Ian0, yes really - feel free to build and test or LT spice it ... if you are able ...

So bad that I can even see the crossover distortion on the trace, even at 1kHz.

 

And full power bandwidth to more than 500Hz! Marvellous!

 

It would be interesting to see the op-amp output - LT Spice is not the hottest on exact opamp models

certainly the real world version has wider freq response and lesser cross over - 22pF across the 120k would assist LT spice.

 

Look how much is the crossover distortion at low levels. It will be massive.

The distortion, poor high frequency response and slew rate and noise are reasons that the LM358 dual and LM324 quad opamps are never used for audio.

I played a video about "What Does Crossover Distortion Sound Like?" and it was "music" played on an electric Geetar through an old vacuum tubes amplifier driving a shrieking speaker. It sounded like normal acid rock noises.

 

I don't even think guitarists like crossover distortion - they always complain about the sound of transistor amplifiers!

 

Here's a 20-Watt Amp in a single To-220 package, for ~5.oo.
https://www.digikey.com/en/products/detail/texas-instruments/LM1875T-LF05/3640532
.
.
.

The great part of that data sheet is on page 6 where it shows the circuit. That is educational!

 

The Specs actually stomp the Specs of the very best High-End commercial Amps You could buy in the '70's.
.
.
.

 

I wonder if I could upgrade my Marantz 22 25 receiver with a pair of those. It seems that the bias control diodes have failed and so the output transistors get rather hot.

 

Look how much is the crossover distortion at low levels. It will be massive.

The distortion, poor high frequency response and slew rate and noise are reasons that the LM358 dual and LM324 quad opamps are never used for audio.

I played a video about "What Does Crossover Distortion Sound Like?" and it was "music" played on an electric Geetar through an old vacuum tubes amplifier driving a shrieking speaker. It sounded like normal acid rock noises.

The worst part is that it is so simple to eliminate most of the crossover distortion with just a few more parts, resistors and diodes. The transformer coupled bass amps with dual secondary driver transformers were the first to see an easy way to get around most of the crossover distortion. And they were simple to servce as well. Then things got complicated.

 

Ah well - if you want the cross-over distortion completely removed - there is a simple addition - see attached

a rail to rail ( output ) op amp would be preferred - but now the LM358 will have sway - even if its open loop gain is a bit down at 20k - you don't have a lot of energy here anyway