Simple audio power amplifier using low cost component's, by SM5UIU.

One thing that most people that are interested in electronics is to build an audio power amplifier. It shouldn't be to hard to find "lot's" of pages on the net that show's different amplifier construtions. Even if your'e not about to build one your self it sure is fun to study all the different solutions that people have came up with.

This amplifier that I present here is just a fun spare time project. I have a home audiosystem that I'm totaly satisfied with so I had no plan's of replacing that one, I did this just to see if I could make one out of scratch and to learn something new.

Well when doing something from the scratch we first have to make a schematic and it sure is nice to have a software with simulation capabilities. After all research and design work we export the diagram to the pcb program and make the layout. All this take's time and that's why this project has taken 3-4 month's? And one of the reasons for that is that I hadn't a suitable cooler, wich I later on found at a ham radio flee market.

Lo level signal is aplied to line in and DC filtered with C1 and R17, R3, C4 form a small input filter.

Q9 and Q10 form a current generator with D2 (led) as a reference, 3 mA / 4 mA. Q1 and Q8 form a differential pair and Q13, Q14 act's as a current mirror that will make sure that we have "exactly" the same current flowing through both transistor's Q1 / Q8, wich will improve the linearity in the amplifier.

Q3 is the high voltage amplifier, 2N5551 & 2N5401 can handle voltage's up to 160 V and when Q3 is turned of it will have a C-E voltage of up to 80V and wise versa with Q10. The same thing goe's for the output transistor's Q4, Q4, Q11 and Q6, Q7, Q12 when it come's to voltage tolerance.

All these transistor's can handle voltage's up to 100V (max). When looking at datasheet's you will notice that when the C-E voltage is at max eg 100V the transistor can't handle much current at all, perhaps 200 mA even though the peak current capabilities could be 25 Amp's. But that shouldn't be any problem beacuse when C-E voltage for Q11 & Q5 is at max eg 80 V "no" current should flow through them and when they are conducting the voltage C-E shouldn't be much higher that 40V.

Q2 is mounted on one of the output transistor's, it's main purpose is to act as a variable "zenerdiode" and R14 is the pot used to set the "zener" voltage and that will also set the current through the output transistors (50-100 mA is a common quisent current value). The voltage over the zener is the same as all P-N, N-P crossovers that we have in the transistor's, wich will be 3 on the positive side and 3 on the negative. If the voltage drop in a transistor would be 0.6 V each the total voltage would be 3.6 V. But since we use darlington transistor's (wich have internal resistor's) and the drop in the output device's is only about 0.5 V the total drop is (of experience) about 2.4 V. The above picture show's how the output signal could look like if we have crossover distorsion (in this case it's +/- 0.6V before the transistor's start to conduct).

The output stage consist's of Q4, Q5 & Q11 on the positive and Q6, Q7 & Q12 on the negative side. R13, R18, R19 & R22 are emitter resistor's that will ensure that we get a somewhat even current distribution in the transistor's. R9 & C6 is just a simple filter that will short any HF oscillations to ground, think it's called a "zobel network"

R5 & R1 along with C2 form the feedback. 18k / 470R + 1 = 39, so 0.1 volt's in, 3.9  volt's out. And at last you can see that I have choosen to have two fuses in each voltage rail, this to protect my switched mode power supply and ofcourse the amp itself. So I have no other overcurrent protection than the fuses, and there is no speaker delay/dcprotection either but that can easily be implemented...

My test's have also show'n that when I drive the amp from the computer's line out I can have full volume on both master and wave output without driving the amp into any hearable distorsion.

Note: No source schematic or pcb layout has been put out on the net.. if your'e interested in building this circuit I could send them in traxmaker / circuitmaker format to you along with the apropriate library files. Email Sam.Yrlund@Swipnet.SE

Datasheet's
1.  TIP35C / 36C
2.  2N5401
3.  2N5551
4.  BDW93C / 94C



The two etched circuitboard's



Front panel layout



Closeup on IN/OUT put terminal's from the outside



Closeup on IN/OUT put terminal's from the inside



Closeup on the power input's from the outside



Closeup of the power terminal's from the inside



One amp channel show'n here



Complete inside view



Speaker's used at the "friday night test run"



Test setup, with 2 DC supplies in parallell. (Did get max about 14 Amp's in 12V wich is far to low, 35-40 Amp's would be more close to the actual current needed but since this only was a test run so.... (It was enough, at least my neighbours thought so)
Later on I tried my "Big 40 A" supply and the fuses in the Amp blew right away...

 

This page has had visitors and was last updated 2002-03-22.