The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> Simulating load-pull and plotting output reflection at the same time
https://designers-guide.org/forum/YaBB.pl?num=1263989873

Message started by DoYouLinux on Jan 20^th, 2010, 4:17am

Title: Simulating load-pull and plotting output reflection at the same time
Post by DoYouLinux on Jan 20^th, 2010, 4:17am

Hi all,

I am wondering about simulating a class A power amplifier in Cadence. In a documentation of doing load-pull test, it suggested that we can use a PortAdapter to plot constant power contours (in Gamma_L plane) and at the same time, we can plot input reflection coefficient (seen by input port) on this Gamma_L plane.

By doing this we can compromise between max output power and input matching quality. However, what about output reflection seen by the 50-ohm load ???

From what I tried, after I designed an impedance transformation network to transform 50-ohm load to ZL,opt for maximum output power suggested by a load-pull test. After testing S22, its magnitude becomes high and the K factor is less than one around the operating frequency (S22 is about -3 dB).

Is it possible to plot constant output power contours and at the same time, plotting output reflection seen by load, so that we can compromise between max. output power and matching quality at the load side ????

Thank you very much again,

DYL

Title: Re: Simulating load-pull and plotting output reflection at the same time
Post by pancho_hideboo on Jan 20^th, 2010, 4:59am

I can't find out any design issue in your post.
Your questions are no more than usage of very specific vendor's tool.
Please post to Simulator's Board not Design's Board.

DoYouLinux wrote on Jan 20^th, 2010, 4:17am:

In a documentation of doing load-pull test,
it suggested that we can use a PortAdapter to plot constant power contours (in Gamma_L plane)
and at the same time, we can plot input reflection coefficient (seen by input port) on this Gamma_L plane.
By doing this we can compromise between max output power and input matching quality.

This is not "input reflection coefficient" of your Power Amplifier.

Is there truely a description of "By doing this we can compromise between max output power and input matching quality" in documentation ?

DoYouLinux wrote on Jan 20^th, 2010, 4:17am:

Is it possible to plot constant output power contours and at the same time,
plotting output reflection seen by load, so that we can compromise between max. output power and matching quality at the load side ????

I think it is not easy if you use Cadence Spectre and "PortAdapter".
Here "PortAdapter" is not lossless.
But if you use other tools such as Agilent ADSsim or GoldenGate, you can easily do.

If you want to judge stability by K-factor under large signal operation, do LSSP or small signal S-parameter analysis subjected to master large signal steady state analysis with sweeping complex value impedance of ports.
Again see http://www.designers-guide.org/Forum/YaBB.pl?num=1239997125

As another method, use interstage impedance probe with small signal analysis subjected to master large signal steady state analysis.
This mehod can be applicable even in Cadence Spectre and "PortAdapter".

http://www.designers-guide.org/Forum/YaBB.pl?num=1049348496
http://www.designers-guide.org/Forum/YaBB.pl?num=1218635777/5#5
http://www.designers-guide.org/Forum/YaBB.pl?num=1240334878/10#10

If your concenrs are stability in linear operation, SP-analysis or PSS/PSP analysis with Ocean function, lsb() might be useful for your purpose. Here "lsb()" is a function for Load Stability Circle.

Note : Even if you use PSS/PSP, you can't consider nonlinear effects caused by complex load impedance, because steady state operation point is determined for fixed resistance value of port, e.g. 50Ohm.

Title: Re: Simulating load-pull and plotting output reflection at the same time
Post by DoYouLinux on Jan 23^rd, 2010, 11:46pm

Hi Pancho Hideboo,

Thank you very much for your explanation. However,

DoYouLinux wrote on Jan 20^th, 2010, 4:17am:

This is not "input reflection coefficient" of your Power Amplifier.

Why is this not the "input reflection coefficient" of the PA ??

Thank you very much again,

DYL

Title: Re: Simulating load-pull and plotting output reflection at the same time
Post by pancho_hideboo on Jan 24^th, 2010, 3:04am

DoYouLinux wrote on Jan 23^rd, 2010, 11:46pm:

Why is this not the "input reflection coefficient" of the PA ??

Do you understand how "PortAdapter" works ?

If you can understand "PortAdapter" correctly, answer is self evident.