Guest
Hi,
I am running a pss simulation on a few transistors which should act as
a PA.
Ive defined 2 ports at the input and output.
Pin is the power out defined in port 1 under Amplitude (dBm), and the
input wave is a sine at 1 frequency. there are no other tones defined
in the port or anywhere else.
The pss simulation is defined as 1 pss 2.4G 3 -40 0 Large Port1
I am sweeping the input power(Pin) out of port 1 over a some range.The
output power of port 1 is at least 5 dBm higher than the input power
and it has spectral components. This only happens if i choose terminal
or port (fixed (R)). The higher output occurs if i choose port (fixed
(R)) but i get a lower output if I choose terminal.
I assumed that it might be modeling spectral regrowth which takes care
of the spectral components but not the power. the power i am
referring to is at the fundamental frequency.
If i change the simulation variable to lets say the width of some
transistor and fix the input power Pin to be constant. Then if I plot
the power out of port 1 vs the width it changes again as a function of
the width when it should be constant. I am really confused on how the
power out of the port is defined and how it can change.
I am guessing that terminal is calculation the input power using 1/2
VxI*. And that fixed R(port) assumes equal impedances for the source
and the input of the PA so the power is something like Pav= V^2/
(4*Rport)? I ask this because ideally i would like to provide pulse
input and i need to be able to accurately measure the power out of the
port.
Secondly I am having a hard time doing a source pull simulation. I am
using the port adapter block, this works fine at the output but I
think that the simulation cant find a solution at a specific value of
theta because its causing the device to break down. When i simulate
this directly using the traffic light it works fine and I get a bad
data point at the angle stated in the log file but when I run a
parameter sweep i cant get any output data.
Will the convergence techniques in the pss convergence post help? I
tried avoiding the point by placing it in the middle of 2 points that
are somewhat far apart but this didnt work.
Any help with this will be greatly appreciated
Thanks
Travis
I am running a pss simulation on a few transistors which should act as
a PA.
Ive defined 2 ports at the input and output.
Pin is the power out defined in port 1 under Amplitude (dBm), and the
input wave is a sine at 1 frequency. there are no other tones defined
in the port or anywhere else.
The pss simulation is defined as 1 pss 2.4G 3 -40 0 Large Port1
I am sweeping the input power(Pin) out of port 1 over a some range.The
output power of port 1 is at least 5 dBm higher than the input power
and it has spectral components. This only happens if i choose terminal
or port (fixed (R)). The higher output occurs if i choose port (fixed
(R)) but i get a lower output if I choose terminal.
I assumed that it might be modeling spectral regrowth which takes care
of the spectral components but not the power. the power i am
referring to is at the fundamental frequency.
If i change the simulation variable to lets say the width of some
transistor and fix the input power Pin to be constant. Then if I plot
the power out of port 1 vs the width it changes again as a function of
the width when it should be constant. I am really confused on how the
power out of the port is defined and how it can change.
I am guessing that terminal is calculation the input power using 1/2
VxI*. And that fixed R(port) assumes equal impedances for the source
and the input of the PA so the power is something like Pav= V^2/
(4*Rport)? I ask this because ideally i would like to provide pulse
input and i need to be able to accurately measure the power out of the
port.
Secondly I am having a hard time doing a source pull simulation. I am
using the port adapter block, this works fine at the output but I
think that the simulation cant find a solution at a specific value of
theta because its causing the device to break down. When i simulate
this directly using the traffic light it works fine and I get a bad
data point at the angle stated in the log file but when I run a
parameter sweep i cant get any output data.
Will the convergence techniques in the pss convergence post help? I
tried avoiding the point by placing it in the middle of 2 points that
are somewhat far apart but this didnt work.
Any help with this will be greatly appreciated
Thanks
Travis