How can I look at an IV Curve and decide where to bias amp for Class A, AB, B?

[ge]

When looking at a IV Curves, how do I know where to set the quiescent drain current for Class A, AB or B? Running AWR, short of plotting the time-domain and looking for clipping?

--- Updated Jan 20, 2023 ---

In the attached, (plagurized from Qorvo),
M2 is defined as Class A,
M3 is class AB
M4 is class B.
Do I "assume" A is half way between cutoff & saturation, B is 0V and how do I quantify AB?

 

[BigBoss]

Yes, you can assume that
A-Class Iceq=Idsq=Midpoint of Icmax and Vce swings between Vknee and 2xVcc
B=Class Iceq=Idsq=0
AB-Class the current is somewhere between these two.(Generally Push-Pull configuration)

These are theoretical values in according with Load-Line principle. Real design values are specified by some other considerations such as Load Pull, Efficiency etc.

 

[ge]

Ok Thanks.

 

[dick_freebird]

Your mention of AWR makes me think this is a RF amplifier.

So your OP and device attributes will be determined by more
than quiescent current and max Pout. You will also (presumably)
care about things like distortion / intercept points which means
keeping the gain and load devices quite linear, and BW which
wants you to run smaller devices, hotter (conflicting with some
of the other care-abouts while helping others).

I think in the end you'll be finding your way through this
multi-walled box by a battery of simulations to optimize.
You might begin with determining which of the Class-es
is most appropriate to your application (modulation schemes
tend to push hard, on this, and new modulation schemes
that enable use of crappy cheap devices are always sought,
and turned into standards).

No mention of Class C, E, F which are more "switching"
RF PA styles (w/ substantial effort in the match / transform).

 

[ge]

I'm learning my way thru a Doherty. Likely I am looking at a Class AB / Class C using CREE devices. Over past years, I've only done Class A, or cheated a little into Class AB. And that was NPN/PNP, LOL. It's amazing the tools we have today (AWR/ADS) compared to 30 years ago.
I'm running Load Pulls right now on AWR.

 

[BigBoss]

We poors, were in "stone age" 30 years ago..
Only rich and institutional companies could access simulation softwares that run on expensive mainframes.

I recommend your to read Steve Cripp's Textbooks (vol-1 and vol-2) and watch youtube videos of Matthew Ozalas on Power Amplifiers

There are other valuable textbooks and App. Notes of different companies and institutions.

 

[FvM]

35 years ago, we had at least SPICE running at IBM PC and Atari ST.