Build an Audio Amplifier

1082

18:29:47

I built it. In both worlds. It works pretty well. Green is the output, blue is the speaker voltage, orange is the audio input.

Here is the list of components needed with some notes on why.

Transistors -> 3904, 3906
Diodes-> 1n4148
Trim pot -> 20k
Normal pot -> 50k
Caps -> 470uF Electrolytic
Aux cable -> Steal it from your friend
Speakers -> 8 ohm 1/2 Watt
Phone -> Audio input
Phone charger -> 5v (put a cap on it)
1 hour to build  -> Get your shit together

Build notes: I didn't need to dc block the audio.  It made zero difference [1]. Needed to cap the noisy 5v source. Tuning the trim pot helped a lot (7.54k). Just connect the PNP base thats connected to the source to ground, amplify about 10  to 50x then check the voltage on the output. If the voltage is too high or low (relative to ground) turn the trim pot till she outputs 0v. (The trim pot is adjusting the DC offset of the amplifier while the normal pot is for amplifying) 

Why two 8 ohm speakers? Because surround sound... or no wait.. stereo? (Obviously 4 ohm speakers would work if you got them.. and yes its single channel)

Sounds like shit? Turn down the amplifier. You can honestly hear the sound fine without any voltage amplification. Turn the volume up on your phone to reduce noise / dc offset from the amp.

Why diodes? The push pull converter is efficient but sounds like someone left the world's largest fan on because of the transition between push and pull.

Alright im ready for the criticism. Lets see what you got

References
[1] Ears

published 8 years ago

hurz

8 years ago

Like the voltage splitter. Its fine for low power. Have you tried to get ride of the 470uF coupling cap at output? Anyway, its for 4 Ohm a little to small. If your splitter and amp is balanced fine, you should be able to use directly ground instead of negative rail plus cap. Keep it up!

GingerKing

8 years ago

I wanted to connect the speaker across the output and ground. My first design didn't have the cap. It totally works too but I found that once the amplifier is turned too high, the output wave form has a noticeable DC offset and drives more current through the speakers for the same smount of audio volume. It's just a personal thing at this point that slightly reduces the amount of current draw. Fun story, My audio test was a ElectroBoom video and he normally blows up electronics. I thought my circuit blew the output 470uF cap that I put in during the test. Huge mistake to play his video. Totally freaked out

hurz

8 years ago

The increase in current is normal, cuz there are no feedback emitter resistors and the biasing is not tuneable (just two diodes). At least glue them at the bjt's to have a thermal coupling.

Lucan01

8 years ago

Maybe @thebugger know something too

drkblader12

8 years ago

Try adding a -don't quote me on this-zobel capacitance network. I.e: in parallel with output a 10 ohm resister followed by a 100n...or so...cap to ground.

drkblader12

8 years ago

Also, try adjusting the circuit with 1-2v @input which is around what an iPhone or CD player might put out.

drkblader12

8 years ago

I'm just self taught, so don't quote me. I just like Audio :)

add comment in editor