There are many possible solutions. Whether any of them are possible using parts available to you, in the time you have to do the job is another matter. If this is for a class project, were you not advised about what parts to use?

 

No, this is not a class project. I am doing it just for personal education purposes.

edit; By the way, what is the best npn-pnp pair for the Ab amplifier?

 

If you just want to make a 1W into 8Ω amplifier, it will be most easily done with a simple amplifier IC using a higher voltage supply. The LM386 is not really man enough for the job though.

Building a discrete class AB circuit bwould be more educational, but some might think it a waste of time nowadays. Again, it would be far easier to do this using 10V or more of power supply, or you would have to do a bridged design. In the end it's your choice as to how much of a job you want to make for yourself.

 

There are many little low power amplifiers asvailable. Most of the new ones are in tiny surface-mount cases and are difficult to solder their microscopic close-together pins.

 

OK, how about this.

Lets take this 8 ohm 0.2 Watt speaker and use this class A amplifier? Dont worry about the constant current source or the 2V source, I know how to implement these things.

 

Your new circuit with the 2N3055 heater does not make sense.
The input from the generator is at 0VDC so the transistor is simply a rectifier. It conducts only when the input is +0.7V or more and is off the remainder of every cycle.
The transistor must have a DC offset voltage of about +3V at the signal generator so that the transistor is always conducting, getting hot and melting the speaker.
The speaker will always have an average of +2DC across it which heats it with 0.5W so it might melt. The DC current will force its cone over to one extreme so it will sound awful.

 

I think you are wrong in this one. What the current source really does is creating a negative voltage in the emmiter VE=-0.3*8=-2.4V which means that even if the source at the base is 0 volts then Vc=5V which is bigger than Vb=0V which is bigger than the Ve=-2.4V so the transistor will conduct all the time because it is in the active region.

In other words, as soon as I have created an emmiter voltage Vbe volts smaller than the lowest possibly base voltage, the transistor conducts.

In addition, the emulator I am using has osciloscope and I have checked the output. It is not a rectifier, it is a full wave amplifier.

 

I think you are wrong in this one. What the current source really does is creating a negative voltage in the emmiter VE=-0.3*8=-2.4V which means that even if the source at the base is 0 volts then Vc=5V which is bigger than Vb=0V which is bigger than the Ve=-2.4V so the transistor will conduct all the time because it is in the active region.

In other words, as soon as I have created an emmiter voltage Vbe volts smaller than the lowest possibly base voltage, the transistor conducts.

In addition, the emulator I am using has osciloscope and I have checked the output. It is not a rectifier, it is a full wave amplifier.

A current source like this is an idealisation. For a practical implementation of such a circuit, a negative supply driving something like a transistor current source would be needed.

You might be better to follow a conventional amplifier circuit, rather than try to re-invent the wheel. Here is a link to a .pdf of some instructions from a kit amplifier - no guarantees that it is error-free, but seems to be quite clearly written. It may be worth a read as a general description, even if the detail of the amplifier design is not quite what you need.

http://kitsrus.com/projects/k48.pdf

 

I think you are wrong in this one. What the current source really does is creating a negative voltage in the emmiter ...

But your schematic does not show a negative supply for the current source, it shows only a positive supply.

Why make a class-A audio amplifier? It wastes a lot of power making heat even when it is not playing.

 

You are simulating the circuit with idealized voltage and current sources and then wondering why you can't built it that way. Avoid these in you simulation if you want to know how the actual circuit will work.

An idealized current source, for example, will generate either a positive or negative voltage as needed to force the required current. No real current source can do that (at least not the simple ones normally used in circuits).

 

But your schematic does not show a negative supply for the current source, it shows only a positive supply.

Why make a class-A audio amplifier? It wastes a lot of power making heat even when it is not playing.

The current source is ideal and it is indeed creating a negative voltage in my circuit, and I am quite sure that it is working properly.

But, the fact is that Adjuster is just right. I tried to implement the current source with real world transistors in the emulator and I lost too many hours creating a bulky, too-many-transistors amplifier.

I will just use the lm386 with a 0.2-0.5 Watt 8ohm speaker. For the record, the reason I need this low watt amplifier is a ta7642 am radio receiver.

I wish i could use it with a tda7000 but I cant find any cheap ones and i cannot afford 5 euro for just one chip.

Anyway, thanx a lot for your help.

 

Ceidas,

I'm not sure what is unclear to you, but this should clear things up a bit...

Power = I*I*R = (V*V)/R.
1 watt = I*I*8 ohms = 0.3536 amps rms or 353.6 mArms

-or-

1 watt = (V*V)/(8 ohms) = 2.828 Vrms

Note also:
1 watt = V*I = (2.828 Vrms)(.3536 Arms) = 1 watt.

To deliver 1 watt to an 8 ohm load you need 2.828 Vrms ON THE LOAD IMPEDANCE, which means that whatever circuit you build and connect the load to, that circuit has to have a low output impedance itself. Note that, if you have a +/-4 Vrms output from any amplifier whose output impedance is 8 ohms, and you connect an 8 ohm load to it, the output voltage is roughly halved.

Your original goal was to drive an 8 ohm speaker with 1 watt of signal power with a +5v supply.

Can you do it? Yes. Easily? Not really.

The best case with a one (ideal) transistor solution is +/-2.5v into the load. This won't work because of the math above.

Opamp - most opamps need more than +/-2.5v (+5v to ground) to even be on. They're just bundles of BJTs/FETs anyway. This won't work either...not to mention 30 mA max. output on most decent opamps.

You need to build a circuit such that the product of the RMS voltage and current equal your 1 watt. Once you do that, you need to match the output impedance of your circuit to the speaker.

+5v is tough to work with, seeing as transistors themselves need 0.7v just to turn on the base-emitter junction, not to mention the headroom you need to reliably operate a BJT, for example. Opamps are worse because they are just bundles of BJTs/FETs anyway. You need to somehow up the voltage of either the output or the supply...

Suggestions:

Use a DC-DC converter to convert +5v supply into +9v (or some other voltage). Your +5v supply will need to be able to source enough current to allow the voltage to be stepped up and still be able to supply your 353 mA per rail (above and below bias point, or, with split supply, up to/down to the limits of the supply).

At the input, making your input signal bigger doesn't help you much. You'd still be limited with your lack of voltage on your power supply.

At the output, an impedance-matching transformer could be used if your input signal to the primary side of the transformer is driven correctly. You could use your class-A 2N3055 stage to drive the primary of a transformer and possibly attain what you're after. Don't forget that the 2N3055 needs driven by a decent-sized signal at the base, also.

Efficiency is a whole other matter. Class A amplifiers are not efficient. Class AB are, at max, about 75% or so. If it were me, I'd up the supply voltage. If I couldn't, I'd use a small audio output transformer like you used to see in RadioShack project kits. Audio output transformers were also popular with tube circuits where, once again, output impedances of tubes needed matched to lower impedance speakers...

Hope this helped.