Hi there!

For a homework project we've been given a class AB amplifier to analyse and solve. The circuit is linked to a Vcc=12V power supply to boost the signal at the npn collecter terminal and to ground at the pnp emitter terminal. However when I model the circuit in LTspice it seems like only the npn transister is contributing to the output while the pnp does nothing at all.
Do I need to connect the pnp to a negative power supply in order for it to work?

Thanks

 

First observation - Q2 needs swapping of collector and emitter. It's incorrectly wired in circuit.

Probably also couple the load [R3] through a capacitor for DC isolation.

I'd also make R1 & R2 an order of magnitude larger.

 

thanks, I've changed those three things and my pnp transister works. But is it normal to have the current values of the pnp transister almost zero compared to the npn?
I know the transisters each should be on the same amount to amplify a sinusoidal signal since it spends as much time negative as it does positive

 

Must be something wrong with your simulation.

I've set up a similar arrangement in my simulation program [SIMetrix] and obtain almost equal quiescent currents in both transistors with zero input signal.

As I increase the signal level the mean current per device increases but the small difference between device currents is still maintained.

 

thanks, I've changed those three things and my pnp transister works. But is it normal to have the current values of the pnp transister almost zero compared to the npn?
I know the transisters each should be on the same amount to amplify a sinusoidal signal since it spends as much time negative as it does positive

I believe there needs to be some resistive feedback from the output back to the input?

As drawn, the NPN has to source about 600 mA into the 10 Ohm resistor load so that extra current flows through it compared to the PNP which has only bias current in it.

 

Did you put a capacitor before the load to block DC?

 

Yup capacitor for the load has been put in , thank you But I cant seem to solve the problem of completely different bias currents in the pnp and npn or have I created the schematic wrong? I don't think I'm meant to add feedback to this circuit either

 

Show as update diagram. Then we can help.

 

Sorry, here's my updated schematic too

 

Diagram looks good.
IcQ1 = 88.96μA and IcQ2 = 89.24μA

And this slightly different in bias currents is caused by a difference in BJT's current gain Hfe (β). But this is not a problem.


Also press Ctrl+R to rotate the component and Ctrl+E - to mirror

 

I agree with Jony130 that the circuit looks fine. Mind you, a 100nF coupling capacitor at the load is rather too small for a 250Hz {?} input frequency. Perhaps a 100uF would be more realistic.

I'm also surprised your simulation is showing a big current imbalance in Q1 & Q2 since (as shown in your attachment) the mean voltage appears to be very close to 6V at the Q1,2 emitters.This would presumably indicate a fairly good balance in the two currents.

 

If I was to hook this up to a speaker modelled as a resistence would I still keep the capacitor at the load? Or would that just take all my dc at the output

 

Yes, you don't want DC flowing through the speaker coil, which could be possible in the event there is no AC input drive signal. If you had split power supplies, say +12v and -12v and the speaker was attached to 'ground' or 0v between the supply rails, then you might not need the coupling capacitor. Of course, even then a failure in the circuit somewhere could result in a DC voltage being applied to the speaker coil.

I edited the reply and change 'yes' to 'no' because when I led my sentence with 'no' it might have sounded like I was saying to omit the capacitor, which is exactly the wrong thing to do. Hopefully the context of my reply made it clear that omitting the capacitor can be a bad thing.

 

No, you don't want DC flowing through the speaker coil, which could be possible in the event there is no AC input drive signal. If you had split power supplies, say +12v and -12v and the speaker was attached to 'ground' or 0v between the supply rails, then you might not need the coupling capacitor. Of course, even then a failure in the circuit somewhere could result in a DC voltage being applied to the speaker coil.

Yes, you need to use an output coupling cap in a working circuit of this type.

To clarify:
If you have a coupling capacitor in the output circuit, DC will not flow into the 10 ohm load with no AC signal, because if 6vdc is applied continuously to the cap from the push-pull transistor stage, the cap will discharge down to zero volts at the load and stay there until an AC signal is supplied.

But, the transistors in your simulation can't actually supply the required current to drive a 10 ohm or 8 ohm speaker load to a useful volume level. You would need additional higher current stages to accomplish this..

 

If theoretically I was to omit the capacitor and the bias conditions for both transistors were then different, what would be the effect on the circuit and its output?
The npn's collector current would be much larger and the ouput current would roughly be 500mA with little or no current going through then pnp, what would this do to a sinusoidal input?

 

Watch this video which may help clear things up for you.

 

Excellent tutorial thatoneguy !
I want some of that test equipment