Building discreet npn multistage transistor amp +-12v psu

Thread Starter

timearley89

Joined Jul 31, 2012
7
I want to build a discreet amplifier from npn transistors. Ive got some n2222s and also some tip3055s, and a dual 12v (+-) power supply. The 3055s are power transistors rated for 90w each, and i have 4. I want to power a 4 ohm speaker across the audio frequency range with 100W. I dont know where to start though. Anyone have any ideas?
 

wayneh

Joined Sep 9, 2010
17,496
I want to build a discreet amplifier from npn transistors. Ive got some n2222s and also some tip3055s, and a dual 12v (+-) power supply. The 3055s are power transistors rated for 90w each, and i have 4. I want to power a 4 ohm speaker across the audio frequency range with 100W. I dont know where to start though. Anyone have any ideas?
You'll need to provide more info. For instance, why are you doing this? Just for fun? Your primary design goal so far is to use NPN transistors. That's an odd starting point. How about details on your power supply and its current rating?

When you say "100W", do you mean rated at 0.1%THD or less, like a hi-fi amp? You don't really need a full 100W continuous do you?
 

Thread Starter

timearley89

Joined Jul 31, 2012
7
Its an amp thats going to power a 15" speaker, its eventually going to go into a home theater system so well realistically say 80W RMS, at less than 0.1% THD. The speaker is a 4 ohm load, and its taking the input of any headphone jack or rca line level output. My power supply(i can build one if i have to but i dont have the parts currently) is an old pc atx supplyrated for 480W with 12v at 16.0A. I have a crude circuit built with 2 n2222s one d3246(from an old car stereo) as a gain stage and 4 tip3055s in parallel as an emitter follower buffer stage, but distortion is fairly high and the heat generated makes no sense in comparison to the 90W rating on the power transistors, i think its a bad design anyway.
 

mcasale

Joined Jul 18, 2011
210
You did not post a schematic, so I assume this amplifier is a Class A.

You are going to cook your components unless you have adequate heat sinks.
 

Thread Starter

timearley89

Joined Jul 31, 2012
7
Yes its a class A, and i have a heatsink, but im just trying to understand how i can arrange the components i have into a functional design, one that is stable and low distortion. I know im limited on the amount of power i can put out simply by the low voltage rails but i want to do as well as i can with it.
 

#12

Joined Nov 30, 2010
18,224
You need to start with math.
80 watts into 4 ohms
PR = E squared
80 X 4 = 320
sqrt 320 = 17.888V
That's the RMS value. To get peak to peak from RMS will cost 2.828 times that voltage. That's 50.6 volts rail to rail if you could build an amplifier that uses absolutely all the voltage available.
 

Audioguru

Joined Dec 20, 2007
11,248
The 60W amplifier is not class-B. Instead it is class-AB like most hifi amplifiers.
When it has a 60V supply then its maximum output swing into an 4 ohm speaker is about 52V p-p which is 18.4V RMS then its output power is 85W.
 

SgtWookie

Joined Jul 17, 2007
22,230
Don't count on the -12v in a computer supply to put out anything over around 1A. You'll very frequently find the -12v regulated by a 7912 fixed linear regulator. Just the positive supply will be usable.

Don't try to connect multiple computer supplies in series, either; the grounds are common with the power cord ground.
 

Thread Starter

timearley89

Joined Jul 31, 2012
7
Im on an android phone and cant upload a schematic for some reason, if you tell me your email ill send you the schematic, but i couldnt get the biasing right on the one i have built right now, there seems to be a phantom 3.6v (roughly) dc signal appearing after the input capacitor at the baseof the first transistor, and i cant figure out where from. Could someone take a look?
 

#12

Joined Nov 30, 2010
18,224
E = IR and P=IE
where E means voltage in "volts", I means current in amps, P means power in watts, and R means resistance in ohms.

These two formulas are the first to learn and are used every day, several times.
 
Last edited:

Thread Starter

timearley89

Joined Jul 31, 2012
7
Im on an android phone and cant upload a schematic for some reason, if you tell me your email ill send you the schematic, but i couldnt get the biasing right on the one i have built right now, there seems to be a phantom 3.6v (roughly) dc signal appearing after the input capacitor at the baseof the first transistor, and i cant figure out where from. Could someone take a look?
 
Top