Maker Pro
Maker Pro

Load-pull limitations in PA design?

B

billcalley

Hi All,

I will be tasked with finding the proper Zin and Zout for a 1W RF
power transistor's maximum Pout, gain, and PAE on a Maury load-pull
station, but something has just occurred to me: won't any results
coming from such a PA load-pull tuning method be relatively
inaccurate? The reason why I say this is that the final (non-linear)
PA circuit, when eventually placed in the RF transmitter itself, will
be seeing a bandpass filter and/or an antenna, both of which mean that
the PA will no longer be seeing the very wideband 50 ohm termination
it saw in the load-pull tuning station. This means that the non-
linear's PA harmonics will be strongly reflecting back off of the
filter's and/or antenna's stopbands and back into the PA's output,
affecting its PAE, stability, gain, etc. Aren't I correct about
this? And if so, how do I address this problem??

Thanks for any help!

-Bill
 
S

Stanislaw Flatto

billcalley said:
Hi All,
This means that the non-
linear's PA harmonics will be strongly reflecting back off of the
filter's and/or antenna's stopbands and back into the PA's output,
affecting its PAE, stability, gain, etc. Aren't I correct about
this? And if so, how do I address this problem??

Thanks for any help!

-Bill
In ANY design, the nicely manicured specs, achieved on bench should be
derated about 50% once you produce more than one unit on production line.
Happens in all designs.
And it is NOT a problem, this IS reality.

HTH

Stanislaw
 
J

joseph2k

billcalley said:
Hi All,

I will be tasked with finding the proper Zin and Zout for a 1W RF
power transistor's maximum Pout, gain, and PAE on a Maury load-pull
station, but something has just occurred to me: won't any results
coming from such a PA load-pull tuning method be relatively
inaccurate? The reason why I say this is that the final (non-linear)
PA circuit, when eventually placed in the RF transmitter itself, will
be seeing a bandpass filter and/or an antenna, both of which mean that
the PA will no longer be seeing the very wideband 50 ohm termination
it saw in the load-pull tuning station. This means that the non-
linear's PA harmonics will be strongly reflecting back off of the
filter's and/or antenna's stopbands and back into the PA's output,
affecting its PAE, stability, gain, etc. Aren't I correct about
this? And if so, how do I address this problem??

Thanks for any help!

-Bill
personally i am much opposed to using non-linear amplifiers without truly
compelling requirements to the otherwise. efficiency is more dependant on
circuit topology to the point that is is unreasonable to be non-linear.
 
J

Joerg

joseph2k said:
billcalley wrote:



personally i am much opposed to using non-linear amplifiers without truly
compelling requirements to the otherwise. efficiency is more dependant on
circuit topology to the point that is is unreasonable to be non-linear.


Power amplifier are pretty much all non-linear, except for classic
single-sideband or TV power amps. But even there tricks such as pulse
modulation had made some inroads. However, in this age of spread
spectrum comms and digital TV (if and when it really comes...) their
days may be numbered.

Just think about it: The big final power amp of an AM transmitter
usually has an efficiency well north of 80%. That wouldn't be possible
if this was a linear amp. These days it's basically a very fast switch
with it's supply voltage modulated via a PWM stage.
 
Top