I will try to make better titles, srry about if the title was mis-leading, it was my very first post so i wasnt aware of it.
But the community here is great and the responses were pouring in.
I have to build a low distortion audio power amplifier for school and i have the schematic already.
I dont get some parts of it, is it wrong to ask for help bc its a school project? I dont want anyone to do the project for me, i def wanna learn on my own but that being said some pointers would def help. Also this goes to everyone, please give your input on this
Thanks again, Appreciate all the help you guys are awesome
Cheers
It's certainly not wrong to ask questions. The thing that is annoying is people wanting someone to do their homework for them.
You are going to need a way to measure the distortion. Traditionally you would use an HP audio oscillator and distortion analyzer. The Wein bridge oscillator was Bill Hewlett's MS project at Stanford. His thesis supervisor suggested the idea which worked so well he encourage Hewlett to manufacture them.
Look for "Max Wein, Mr. Hewlett and a Rainy Sunday Afternoon" by Jim Williams. I think you'll find that is a nice follow on project.
In the mean time, you are going to need a way to measure the distortion. MATLAB or Octave and a sound card are the best option. If it's not already obvious, you will need to generate a WAV file with a 1 KHz sine wave and play it through the sound card line out while recording it on the line in using a cable from output to input. That will allow you to determine the errors in the sound card input and output.
Then repeat the measurement with your amplifier in the circuit with a suitable attenuation pad. Load the output into MATLAB, compute the amplitude spectrum, normalize by the amplitude spectrum with just the cable and then sum up the harmonics and divide by the fundamental.
If you've already had a course in linear systems this should be trivial. If not post or send me a PM and I'll explain the fine print details.
With regard to the start of this thread, as you are an EE student, you should be able to arrange lab time to test personal projects for which you don't have the test gear needed. That makes building test gear a lot easier than it was for me. My first degree was English lit, my second was geology, so I couldn't get in the EE labs and only had a recurrent sweep 5 MHz Heathkit scope I picked up at yard sale for $75. But it was enough to build and test a linear 5 V supply for TTL and repair a 12 V supply I bought for $1.
For a signal source I used a quasi square wave generator using an op amp. I don't think I ever added a TTL divider to make it truly square. I had great fun building a 120 dB 3 transistor DC coupled amplifier. At the time I was obsessed with building a 40 M xcvr to fit in a cigarette pack (this was 30+ years ago when it was hard to do). So the 120 dB amp was built on about 1" square on perfboard. And it oscillated like mad at 200-300 KHz. I studied the circuit and decided that the problem was the electrolytic bypass capacitors which have significant inductance. So I added some small ceramic caps in parallel and after that design change it worked just fine. I was quite proud of myself for figuring that out as I'd never taken an electronics course. And still haven't. The closest I got was listening to lectures as a break from looking through a microscope all day at thin slices of rock for my thesis project.
120 dB is a *lot* of gain. With a small speaker hooked up the output I could listen to a nearby FM station just by holding a finger close to the input terminals by means of slope detection.
FWIW you could build a quite high performance audio test bench around a Raspberry Pi or similar. You'd have to write custom software, but it would not be hard as the difficult stuff like the FFT code is widely available. There are easy ways to increase the dynamic range of the measurements far beyond what the ADC in the Pi will do using DSP.
Have Fun!
And if you have a masochistic streak, try to build a sub-nanosecond rise time pulse generator for testing scopes using a 10 MHz clock module and a fast logic driver for the output. It's a lot harder than it might sound. But there are some really good suggestions for the output in various forum threads.