A Few Basic Questions About Power Amplifier Theory - RF Cafe
Post subject: A few basic questions about Power Amplifier
theory Posted: Tue Apr 18, 2006 11:05 pm
Joined: Tue Apr 18, 2006 10:40 pm
i am starting
a class A PA design project and i had a few basic unanswered questions.if
someone can answer them it will be nice.
1)why does saturation
occur? i know there might be many reasons but is it also because
that the transistor transconductance,gm, starts to reduce and then
reaches some steady level? if so, why does gm decreases and then
reaches some certain level? but if what i said is true then even
at high power, as far as i understand, the bias circuit is still
providing the same base current and thus same Ic so why would gm
be changing since gm=Ic/Vt?
2)this could well be explained
by above answer--i biased for class A and this bias is for certain
output power that i want and this requires some certain intput power,
Pin1. what happens when the input power is less than Pin1 as far
as dc biasing in concerned? what happens when input power is at
a high level? what i am really interested is in how the bias voltage
Vbe and current Ic change? if they increase (i don't know why they
would) then gm should also increase and if that's truly the case
then the saturation is not occuring because of gm (question 1)?
3)what is the best book for power amplifier theory/design. thanks
Post subject: RF PA DesignPosted:
Wed Apr 19, 2006 7:06 pm
Feb 22, 2006 3:51 pm
Saturation in an RF
amplifier occurs because the collector or drain voltage swing is
limited - it can't go negative. (In fact, it can't go below 0.7V
for a bipolar transistor if you don't want excess charge in the
base region - which causes all sorts of bad results). So the downward
swing is from Vsupply to about 0; that means that the upward swing
can be no more than 2 x Vsupply. Limited voltage = limited (saturated)
Please observe that this is not the same as bipolar
transistor saturation, which occurs when Vce<Vbe.
A service, the transistor (BJT or MOSFET, doesn't matter) never
has zero collector/drain current and never has zero collector/drain
voltage. Likewise, it never exceeds 2 x Ic center or 2 x Vcc.
Regarding the book recommendation: What power levels are you
Post subject: Posted: Wed Apr 19, 2006 7:47 pm
Joined: Tue Apr 18, 2006 10:40 pm
hi fred...thanks for the reply. it makes sense, you are saying
that as the input power is increased the output voltage swing can't
increase accordingly because of clipping, meaning that after optimum
power output levels the voltage swing is stuck at the level but
input keeps increasing so the saturation occurs. But, another question:
if I keep increasing the input power won't the Vbe become large
and evetually Vce<Vbe and that the transistor is in saturation
and the gain drops to ~unity, right?
as far as book is concerned
i am looking for a good book on overall PA design...power level
doesn't matter. but if you want to know i am looking at moderate
power levels not high levels. Levels needed in wireless communication
Post subject: Posted:
Fri Apr 21, 2006 8:46 am
Jun 27, 2005 2:02 pm
i biased for class A and this bias is for certain output power that
i want and this requires some certain intput power, Pin1. what happens
when the input power is less than Pin1 as far as dc biasing in concerned?
what happens when input power is at a high level?
answer your question: If the input power is less than Pin1 then
your amplifier will have lower efficiency. Class A by definition
is the most inefficient Class of operation, yet the most linear
one. When the input power is at a higher level the efficiency will
Post subject: Posted: Tue Jun
13, 2006 10:11 pm
Joined: Tue Mar 15,
2005 11:43 pm
Hi Jabb :
Regarding the good textbooks for power amplifier design, you
may want to try the followings :
1. Gonzalez Guilermo, Microwave
Transistor Amplifiers Analysis and Design, New Jersey: Prentice-Hall
Inc, 1997, 2nd edition.
2. S.Cripps, RF Power Amplifiers for
Wireless Communications, Artech House, 1999.
3. S.Cripps, Advanced
Techniques in RF Power Amplifier Design, London: Artech House, 2002.
Hope it helps.