Power Added Efficiency

[No one]

H
Please decribe the phrase: Power Added Efficiency or PAE.
Introduce a good descriptive book about it.
Nooooooooooooooooooooooooooooo
Oneeeeeeeeeeeeeeeeeeeeeeeeee

 

[GlamZ]

The efficiency of a PA is quoted as it's output power vs the DC power necessary to produce that output. IE. a 1 watt PA that draws 2 watts of DC power is 50% efficient.

... However using this figure a PA could require that you put 0.99999W in to produce the 1W output at 50% efficiency drawing 1.99998W to do it. This would not be a usefull or efficient PA!!

PAE takes this into account by using the added power not output power
(Pout - Pin)/Pdc. so for the above PA the efficiency is 50% and sounds quite good, but the PAE gives a (more honest) figure of 5*e-6%!

( This can be also written as PAE = efficiency*(1-1/Gain). )

Any general RF book or PA book will have this in it, but it's a simple concept and not worth buying a book just for that, Check out the ebooks section for some general texts.

Regards.

 

[mckinson]

Hello Glam:


Do you know how to improve the PAE ?
What is relationship between the PAE and PA saturantion ?
Best regards !

 

[GlamZ]

>Do you know how to improve the PAE ?

It depends on what you need to do, There is a trade off between linearity and Efficiency!!
Generally speaking you will need a higher gain device, If you are not using an system where linearity is important for suppressing IMD. IE. FSK not ASK you can switch to a more efficient PA class such as C or D rather than A, or rather than having one PA stage have the PA split over 2 or 3 stages with different classes of operations to produce an average PAE figure. Adaptive Bias control is another way to go.
If it's really a problem you need to consider an approach such as a Doherty or EER configuration, but these tend to have there own drawbacks.

I'd suggest reading 'any' book by Steve Cripps (there should be a few under the ebooks section)

>What is relationship between the PAE and PA saturantion ?
PAE will fall as the PA saturates, because to saturate the PA you keep increasing the power input, whilst the DC power and Power output remains constant (because the output has saturated). Think about it the Gain effectively falls whilst the DC drawn remains constant so the PAE figure must fall. The DC power and the output power effectively become constants in the previous formula, so it will vary from one PA design to another PA design.

Hope this helps.

 

[Wilson_yu_chen]

It's seems you are talking 1W power output application. If for GSM application, then
, little complicated to reach better efficiency. Because with higher efficiency, you would suffer less power output as PA's working on saturation mode in high power.
We usually painfully get 3-5% efficiency with minimum power requirement and power sideband performance trade-off by playing load-pull.