Post subject: Whats best for amplifier efficiency...
Large V or small V ?
Unread postPosted: Wed Oct 06, 2004 11:42
If I run an amplifier with a large DC supply rail (lets
say 12 Volts) as opposed to a small supply rail (lets say 3 Volts)
then for the same RF power output the currents flowing around are
going to very much smaller in my 12Volt system than if I had operated
with the 3 volt rail. With this in mind I then think (maybe incorrectly)
that the losses in my high voltage system must be much lower than
the 3 volt system (I am assuming I squared R loss dominates). I
am not sure if the assumption that I squared R loss is the dominant
mechanism generally but I would imagine that it is for most HF,VHF
and UHF to say 1GHz amplifiers ? Now if this is all sound thinking........why
are we seeing lots more lower voltage transistors and less and less
of the higher voltage ones?
Of course there is the argument
of where you get a large supply rail from if limited to small voltage
batteries or a low supply rail but then you can always add a switch
mode step up converter!
Anyone care to comment on all this?
Wed Oct 06, 2004 12:46 pm
From what I
understood you are dealing with a low power amplifier eg MMIC or
Transistor. Amplifier efficiency can be calculated by the ratio:
Pac/Pdc. The amplifier has to draw a constant current which is the
bias current. You can set this current by a voltage drop resistor,
which can be calculated by: R=(Vrail-Vd)/Ibias, where Vd is the
device voltage. If you use a high rail voltage then the overall
circuit effiency is degraded due to the power dissipated on the
resistor, however the efficiency of the amplifier will remain the
same. if you draw more output power from the amplifier then you
will increase the efficiency.
I hope this helps,
Thu Oct 07, 2004 8:25 am
Thats all fine but I am thinking
more in terms of class B or C. In these cases you dont have a bias
resistor in the emitter! Since my last post the only other thing
I can think that would be a reason to run a lower supply voltage
is the benefit to impedance matching. From the basic theory for
class C amplfiers with a collector choke the required impedance
at the transistor collector is approximately VCC^2/(2xPload) So
from this equation you can adjust VCC to get your required impedance
as near as possible to 50 ohms (assuming a 50 ohm system). You then
do not need to impedance match (apart from a bit of reactive cancellation).
Apart from the above argument I still do not see why we are
seeing so many low voltage transistors around!