Ok so I tried my setup with a newly recharged toy car 7.2V 6800mAh battery and ran the unit for about 5 minutes. The sound is nice at all volumes but I noticed the tip42c transistors were getting warm but the tip41c transistors remained cool. Looks like I'm going to have to re-attach the heatsinks again.

 

When you change battery voltage you need to readjust the bias point on the driver transistor. Adjust the bias resistors until you get half the supply voltage at the DC output.

 

How about this one?

View attachment 274516

Isn't that a lower current model without bootstrapping? My pull-up power NPN base is fitted with 330 ohm resistors and why pull up the first NPN base input to VCC with a resistor? I'm getting power pulling it up from the emitters from the power transistors via a large resistor. Sorry I'm trying to get the best power off a small voltage without blowing anything up.

 

If you bought real transistors from a genuine electronic parts distributor then they should heat similarly if you have the biasing correct for a symmetrical voltage swing. Cheap fake transistors from China could have very poor matching.

With a power supply of only 7.2V then the output is about 6Vp-p producing only 0.56W into 8 ohms with some distortion and total heating of another 0.56W (0.28W for each output transistor) if you play a continuous tone, not music or speech that has a lower average output and heating power. A heatsink is needed if each output transistor heating exceeds about 1.5W.

Maybe your method of construction (solderless breadboard?) is causing the circuit to oscillate at an inaudible very high frequency that produces heating.

 

You are messing around with this circuit without fully understanding the function of each component in the circuit.

This thread started out with the objective of calculating the output power.
You cannot get more power out of it without sacrificing sound quality. If you want more power output you have to be prepared to make changes to the design.

If you would like an explanation of what each component does I would be happy to show you.

 

If you bought real transistors from a genuine electronic parts distributor then they should heat similarly if you have the biasing correct for a symmetrical voltage swing. Cheap fake transistors from China could have very poor matching.

Again, I'm on a budget. Despite the warmth, things are looking good even though I set the resistor ratio to 56k/100k instead of 56k/160k (where 1st number is resistor from NPN base to ground and 2nd number is feedback resistor).

Maybe your method of construction (solderless breadboard?) is causing the circuit to oscillate at an inaudible very high frequency that produces heating.

My construction is on a copper clad PCB with traces about 50mils wide and the lengths are very short. The speakers however are connected to the board via twisted 4 inch wires. Same goes with the battery.

I wonder if it could be because I cranked up the source music to maximum volume possible because I didn't really notice warmth before that.

I also set the bias the way I did because I verified it with LTspice and because I didn't want sound to be distorted when the battery level drops by a mere 5%.

 

I am not sure which circuit diagram you are using.
Replace one of the bias resistors with 75% of its value and add a trimmer resistor in series with 50% value.
This will give you an adjustment range of ±25%.