ClassA question for the enlightened...

[Bernhard]

Hi,

if an amplifier is classA and not single ended, how much idle current is needed ?

Must the idle current trough each transistor be the same or higher then the current through the speaker at all times ?

Or ???

 

[nobody special]

a class A p/p amp can put out 2X the idle current before leaving class A.

 

[Bernhard]

p/p ? push/pull ?

 

[Bjerre]

Hi Bernhard

The answer to your question is yes. The idle current must be larger than the largest current ever to go through the load. Said in another way, the idle current must be large enough for the output tubes/transistors newer to turn off.

And yes, p/p is push-pull !

 

[nobody special]

p/p= push/pull

 

[Bernhard]

Ok.

I want to built a classA amp, where the idle current increases with the signal.

Does that make sense ?

Some kind of second amp inside the amp which lets the current be high on high signals.

 

[nobody special]

Yes, it makes sense. It has been done before. I can't remember what it was called (there are many here much more knowledgable about that).
What kind of output devices are you using? If you're using MOSFETs, it isn't a good idea, because the low-level resolution will suffer.
Steve

 

[Bernhard]

What kind of output devices are you using?

Its in http://www.diyaudio.com/forums/showthread.php?s=&threadid=11205&perpage=15&pagenumber=32

I call it madness circuit.

 

[Pan]

a class A p/p amp can put out 2X the idle current before leaving class A.

Should be 1x idle right?

Or where do you get the extra current from?

/Peter

 

[nobody special]

Should be 1x idle right?

It says 2X idle in the Pass A75 article. That's where I got my info... I think it's correct. Maybe someone can explain why that is. I know I simulated it before, and it made sense, but I can't remember just what happens to explain it better. Sorry.

 

[Christer]

I was about to write an explanation, but perhaps a picture
explains it better. Look at the third set of pictures in
http://perso.wanadoo.fr/jm.plantefeve/mont.html

All text is in French, but I guess the picture is self-explanatory.

 

[Bernhard]

Maybe you can write some text beside the piture

 

[Pan]

That 3rd picture shows that the current out to the load can never be greater than the idle or the amp leaves class A territory.

1x idle it is.

/Peter

 

[Circlotron]

a class A p/p amp can put out 2X the idle current before leaving class A.

I think so too. Imagine the upper and lower transistor are both sitting at 5 amps. Then the amp starts to pull up the load by 10 amps. This means the upper transistor current will go from 5 to 10 amps and the lower transistor will go from 5 to 0 amps. 0 amps is the point it would move out of class A if you went further. The upper transistor takes on 5 extra amps while at the same time the lower one relieves the upper one of 5 amps. Add positive five and subtract negative five. Therefore you can sourced or sink double the quiescent current and stay in class A, IMHO.

 

[Hoffmeyer]

This means the upper transistor current will go from 5 to 10 amps and the lower transistor will go from 5 to 0 amps. 0 amps is the point it would move out of class A if you went further.

This just shows that you'll need more than 1 x idle to stay in classA, since, in the above example, the lower transistor will be in a turned off mode, just outside classA.
With 2 x idle you'll be safely in classA, even if the load is less than expected, wich it often is

 

[Bernhard]

So what is to avoid, is that one transistor is without current, right ?

 

[Pan]

Take a lok at those pictures in that link again.

Let´s say in a class A PP amp we have 1 amp idle. That means both devices are conducting 1 amp. If the PS voltage allows you can get "any" current out to the load but when the load receive 1 amp or more, one of the devices will be off = class AB.

Don´t forgett we have +/- supply and when one device start to close, the other one have to open up to stay at 1 amp "idle".

It´s not like one device closes, the other opens up and the current rises. Think at the two devices as shunts that deals with the idle current.

The idea with 2x idle is not correct.

A class a amp can only output the idle, then it enters AB.

1x idle it is.

/Peter

 

[Pan]

Bernhard,

yes, as soon as one transistor does not conduct any current, then you´r in AB and not class A anymore.

As transistors turn on/off there is a spike and switching distortion occur due to the gross non-linearitys at the turn off point (and even close to it).

/Peter

 

[wuffwaff]

Pan,

somehow you must take a look at that picture again. A push/pull amp idling at 1A can put 2A through the load before the lower transistor switches off.

This means that the peak current of an class A push/pull amp is TWICE the idle current giving 50% of max. efficiency.

With a single ended amp this would be 25%


william

 

[Hoffmeyer]

Pan,

somehow you must take a look at that picture again. A push/pull amp idling at 1A can put 2A through the load before the lower transistor switches off.

william

Lets separate theory from reallity:

Theoreticly William is right, but the conditions there are ideal, unlike reallity, and if you're making a class A amp, you do not want any transistor to get even close to switching off, since no transistor is linear, so how about a compromise here: 1½ x idle, to make some headroom for odd loads