The 1N914 and 1N4148 diodes are a good match to the 2N3904 and 2N3906 transistors to reduce or eliminate crossover distortion. The current in the diodes creates the amount of voltage across them. The current is adjusted by changeing the values of R2 and R4 in Junior's circuit.

Junior, you have the volume control connected backwards so it shorts the output of the generator when it is turned down.
The generator should feed the top terminal of the volume control and the slider should feed the input of the amplifier.

Short together the bases of the output transistors (then the diodes are removed) to see the severe crossover distortion when the transistors operate in class-B.

 

Audioguru is quite correct about the volume control being backwards, which is my fault. I'd thrown the circuit together in a hurry a couple of months ago, and didn't do a good job of bench-checking it. Tsk, tsk!

See the attached image for the correction.

 

how about if i change my transistors to 2n2222 and 2n2907? does it make any change?

p.s. i really want to design my own.. but i dont know how to calculate.

 

I made a new ckt of a class AB pushpull amplifier..

but look at my ckt. i attached

My amplifier tends to attenuate.. pls help..

 

Well, the 2N2907/2N2222 don't have as much gain as the 2N3904/2N3906 pair.

Try using 2N4401/2N4403 pairs.

R3 and R4 - why are you using them?

You need an amplifier to drive the bases of the transistors, otherwise your input will be very low impedance. While your simulated signal generator will handle that (as it can supply infinite current) you will find that a real-world signal source will not be able to drive the circuit very well.

Try putting a 10k resistor in series at node 9, and you'll see what I mean.

 

its still the same sir..

 

I'm afraid that you aren't understanding what the purpose of the first stage of the amplifier is.

Remember in the original circuit, there was an NPN transistor on the cathode of D2, instead of your R2?

Your R2 is a fixed 1k Ohm resistor. However, the NPN transistor acts like a variable resistor, controlled by the input signal.
The NPN transistor caused R1 to pass a variable amount of current through D1 and D2, which translates to varying voltage levels on the bases of Q1 and Q2.

You removed the first stage of the amplifier, which had a gain of roughly 220. Why are you surprised that your version does not work?

 

Remember the two diodes in the base circuit? Those are there to keep the transistors right at the edge of being biased on. If you're seeing crossover distortion, try changing the diodes to a different model.

The type I used, 1N4001, are power rectification diodes, and have relatively slow response times. You might try using a newer version, like perhaps 1N5817, which is a low forward voltage Schottky diode. You will need to use more of these in series, because they have a lower voltage drop across them.

For example, at 100mA, a 1N4001 will have a Vf of approximately 0.75 (Fairchild Semiconductor datasheet, 1N4001-1N4007, Figure 2, Vf Characteristics).
At the same 100mA current, a 1N5817 will have a Vf of 0.32 (ON Semiconductor datasheet, 1N5817 - 1N5819, Electrical Characteristics, Max Instantaneous Vf).

Try replacing the two 1N4001 diodes with four, then five 1N5817 diodes in series. See what happens to crossover distortion. With four, it should be much worse. With five, it should be much better - except your current will increase due to the transistors being biased on a bit more.

You could "fine tune" it by adjusting the current flowing through the diodes. Review the datasheets to see the Vf characteristics. As you will see, the Vf is not cast in concrete; it varies as a function of current.

If the Vf of the diodes is perfectly matched to the Vf of the emitter circuits of the output transistors, the crossover distortion will be minimal along with minimal current usage. But in the "real world", no two transistors are identical (even if they have the same markings) - so Your Mileage May Vary.

hi can you help me... looking also for amplifier class ab... i will use TIP42A transistor, the minimum transistor to be used is 4.. the output must be 30 watts rms output

http://www.datasheetcatalog.com/datasheets_pdf/T/I/P/4/TIP42A.shtml