jason9

modified 8 years ago

Class-A Amplifier

368

05:41:32

Close the switch to activate.

It has about 1-2mV of second harmonic distortion at full power (measured at 317-318Hz), and much less distortion with other harmonics. This means that it has a THD (Total Harmonic Distortion) of around 0.003%. This is not THD+N (Total Harmonic Distortion + Noise), and therefore does not include any thermal noise or any other kind of noise.

Note: I changed the amp by pushing it further into class-A mode, so the distortion might be a bit lower than what's said above. I also measured the efficiency to be 41% in its current version.

published 8 years ago

hurz

8 years ago

This is still a mix of A and B

jason9

8 years ago

Would it help if the 1.2kOhm resistor was replaced with a 1.1kOhm resistor? I believe that the transistors are always on, although they do go into their nonlinear region at the peaks of the wave.

hurz

8 years ago

Two options, replace one power stage e.g. the upper two transistors to a constant current stage or replace it with a resistor. But take it away from the controlling loop.

thebugger

8 years ago

Yeah hurz is right, even a simple modification like this, would work better http://everycircuit.com/circuit/5560997186371584 I personally recommend a bootstrapped emitter follower upper arm, but a straight on constant current source would also work.

hurz

8 years ago

The voltage adaptation stage is also not needed. It makes only sense to set the idle current for a push pull topology.

jason9

8 years ago

Why wouldn't a push pull output stage be good for a class-A amp?

jason9

8 years ago

I measured the constant current source type topology to have a theoretical maximum efficiency of only 25%, while the push pull type topology to have a maximum theoretical efficiency of 50% (assuming the transistors are perfect voltage controlled current sources). Why should I use the constant current source topology?

thebugger

8 years ago

A maximum theoretical efficiency of 50% for a class A amplifier is only with a transformer coupled output.

hurz

8 years ago

Its a class A amp by definition if its pulling against a load as resistor or a current source. While the second option is already a grey field in definition and its up to you if you see it still as pure class A!

jason9

8 years ago

My personal preference is that a push-pull amplifier is a pure class-A amplifier as long as neither of its transistors ever turn off. I guess this amp is actually kind of grey for me since the transistors do go deep into the nonlinear region, nearly turning off. This can be fixed by making the 1.2kOhm resistor 1.1kOhms, but I think I'll just replace it with a 2.2kOhm potentiometer.

hurz

8 years ago

A push pull stage is clearly NOT pure class A.

jason9

8 years ago

Well, it still has very high fidelity, and eliminates almost all even number harmonics, as well as having a class-A type efficiency. As far as I can tell it's just as good if not better than a class-A amp because of the elimination of the even number harmonics. I already said, my personal preference is that this is pure class-A. You just said: "its up to you if you see it still as pure class A!"

hurz

8 years ago

Most of this world gives a shit on your personal preferences.

hurz

8 years ago

I talked about "resistor or constant current"'. Everrhing else is clearly a class b or AB.

jason9

8 years ago

Did you even see what I just quoted you on?

hurz

8 years ago

Have you understand what a class A amp is. Your personal preferences are irrelevant if you make them public. Back to my first comment, a class AB is not a class A amp.

hurz

8 years ago

Removed the push part and just left over the pull stage, its a class A amplifier. http://everycircuit.com/circuit/6461127842332672

jason9

8 years ago

But it has terrible efficiency. I could go with a class-A amplifier that uses transformers to get a more reasonable efficiency, but audio transformers can't operate well across the entire audio range, and I don't know how to limit their idle current when using a differential amplifier to remove distortion.

thebugger

8 years ago

Audio transformers are very expensive, although they can be made to operate throughout the audio spectrum and above. Some transformers work up to 100kHz pretty well, but the linearity is an issue with them

jason9

8 years ago

How do you control the idle current when using a differential amplifier to help remove distortion?

thebugger

8 years ago

The example hurz gave you is very basic, in reality, a more complex current source is typically used, one which has such a control mechanism.

jason9

8 years ago

Yeah, but I want to know how I to do it easily in EC.

thebugger

8 years ago

There are numerous topologies of such a device, look em up online

jason9

8 years ago

Ok.

thebugger

8 years ago

You can try a bootstrapped emitter follower. It's easier to implement

jason9

8 years ago

Doesn't that make a constant current source as would be used in a transformer-less design?

thebugger

8 years ago

In a transformer coupled amp the transformer itself acts as a CCS by default.

jason9

8 years ago

Actually, in idle it just acts as a very low resistance resistor making it so that the differential amp tries to pull down the voltage but the resistor tries to keep it up high and the end result is either no current or extreme current. How do I keep the idle current under control?

thebugger

8 years ago

Yes its reactance to current change, makes it a good constant current source, but only for AC. That's impedance. It's low resistance to DC makes it an unsuitable CCS.

jason9

8 years ago

Ok, but where does the bootstrapped emitter-follower come in?

add comment in editor