AnsweredAssumed Answered

Setup arbitrary frequency dependent impedance on the network

Question asked by xuzhu on Apr 23, 2009
Latest reply on Apr 27, 2009 by xuzhu

Like • Show 0 Likes0
Comment • 5

I am evaluating a special 2-port matching network device over a wide frequency span. The termination impedance on one side of the device is 50ohm, and on the other side is frequency dependent, e.g. antenna impedance. I can use load-pull tuner to mimic this frequency dependent load but I have to change the tuner’s setting for each frequency point. It is a slow process to scan the entire spectrum and every time I adjust the matching network, I have to repeat the whole process before I can know if I tuned into the right direction. So I am wondering if the fixture simulator of the network analyzer can help me on this. Right now, I have an E5071B in the lab, the Port Zconvertions only allows me to put in a fixed impedance value for each port. It doesn’t allow me to load the S-parameter file or impedance data file. I am wondering if there is way to load the S-parameter file into the port of the network analyzer. I notice that De-Embedding and Port-matching functions can load user defined S-parameter files into the network analyzer. Does this require special VBA programs or firmware change?

Xu (972-761-7849)

No one else has this question

Outcomes

RayH Apr 23, 2009 12:07 PM

The fixture simulator "port matching" function will allow you the EMBED an s2p file into your measurements. As long as you have the proper data file, this is the function you need.

This is part of the analzyers fixture simulator functions. It is already in the firmware of every E5071C, and in E5071B with revision 3.50 or greater. Most E5071B's will have this installed already. So, you probably do not have to update or use a .vba

Actions

xuzhu Apr 23, 2009 12:51 PM

Ray,
What I need is S21 measurement over the variable impedance. I looked into user manual, http://ena.tm.agilent.com/e5071c/manual ... ircuit.htm (E5071C and E5071B has the same illustration), the S parameter measurement is performed on calibrated 50ohm terminal. Then here is my understanding: when I put the S parameter of match network into the ENA, the measured S parameter changed but it is still referred to 50ohm terminal, not the variable impedance I wanted. Am I right here?
Xu

Actions

odanzy

Apr 24, 2009 3:55 AM

You are basically looking for a software implementation of a frequency dependent impedance tuner. You could use either method, but neither would be a simple, straight forward process. The Port Z Conversion method allows you to mathematically change port impedance and the data is massaged to show the effects of the new impedance, even though the calibration was completed in 50 Ohms. You would still have to take a new sweep for each data point, but this could be automated pretty easily. With the Port Matching functionality you could create a touchstone file or mimic a simple circuit that would cause the frequency dependent impedance change that you are looking for which took into account the 50 Ohm termination into the test port. There is not inherent functionality that would put a frequency dependent slope in the Port Z conversion routine.

Converting the Port Impedance of the Measurement Result (from ENA WebHelp):
http://ena.tm.agilent.com/e5071c/manual ... result.htm

Actions

Dr_joel

Apr 24, 2009 8:43 AM

The port impedance conversion routine is, essentially, embedding a transformer as an S2P file. If you change the port impedance to 200 ohms, we embedd a 2:1 transformer (Z goes as turns squared) s2p file, under-the-covers so to speak. And then we change the marker annotation by the transformer value too, in the Smith chart display. But Mag/Phase of the S-parameters are modified only by the s2P file. So, for you to create a custom impedance profile, you simply simulate a transformer with a different turns ratio for each point. Embedd this S2P file and you are done if you just want S-parameters, but if you want to see impedance (Smith chart) then you need multiply the marker impedance value by the impedance that you have.

Actually, for complex impedance, it's slightly more tricky, as we have to transform the real impedance then add a reactive element to create the imaginary part.

Or just read the 50 ohm data into ADS and transform it there.

Actions

xuzhu Apr 27, 2009 2:32 PM

Thanks all for helpful information. I really wish the network analyzer can load arbitrary port impedance into the port in the future, therefore I can do testing and tuning in real time as with 50ohm system measurement. Either loading measurement data into ADS then doing impedance conversion or doing point by point PortZ conversion is kind of slow process for my application.

Thanks so much for all you guys' help.

Actions

5 Replies

RayH Apr 23, 2009 12:07 PM
Mark Correct
Correct Answer
The fixture simulator "port matching" function will allow you the EMBED an s2p file into your measurements. As long as you have the proper data file, this is the function you need.

This is part of the analzyers fixture simulator functions. It is already in the firmware of every E5071C, and in E5071B with revision 3.50 or greater. Most E5071B's will have this installed already. So, you probably do not have to update or use a .vba
Like • Show 0 Likes0
Actions
xuzhu Apr 23, 2009 12:51 PM
Mark Correct
Correct Answer
Ray,
What I need is S21 measurement over the variable impedance. I looked into user manual, http://ena.tm.agilent.com/e5071c/manual ... ircuit.htm (E5071C and E5071B has the same illustration), the S parameter measurement is performed on calibrated 50ohm terminal. Then here is my understanding: when I put the S parameter of match network into the ENA, the measured S parameter changed but it is still referred to 50ohm terminal, not the variable impedance I wanted. Am I right here?
Xu
Like • Show 0 Likes0
Actions
odanzy Apr 24, 2009 3:55 AM
Mark Correct
Correct Answer
You are basically looking for a software implementation of a frequency dependent impedance tuner. You could use either method, but neither would be a simple, straight forward process. The Port Z Conversion method allows you to mathematically change port impedance and the data is massaged to show the effects of the new impedance, even though the calibration was completed in 50 Ohms. You would still have to take a new sweep for each data point, but this could be automated pretty easily. With the Port Matching functionality you could create a touchstone file or mimic a simple circuit that would cause the frequency dependent impedance change that you are looking for which took into account the 50 Ohm termination into the test port. There is not inherent functionality that would put a frequency dependent slope in the Port Z conversion routine.

Converting the Port Impedance of the Measurement Result (from ENA WebHelp):
http://ena.tm.agilent.com/e5071c/manual ... result.htm
Like • Show 0 Likes0
Actions
Dr_joel Apr 24, 2009 8:43 AM
Mark Correct
Correct Answer
The port impedance conversion routine is, essentially, embedding a transformer as an S2P file. If you change the port impedance to 200 ohms, we embedd a 2:1 transformer (Z goes as turns squared) s2p file, under-the-covers so to speak. And then we change the marker annotation by the transformer value too, in the Smith chart display. But Mag/Phase of the S-parameters are modified only by the s2P file. So, for you to create a custom impedance profile, you simply simulate a transformer with a different turns ratio for each point. Embedd this S2P file and you are done if you just want S-parameters, but if you want to see impedance (Smith chart) then you need multiply the marker impedance value by the impedance that you have.

Actually, for complex impedance, it's slightly more tricky, as we have to transform the real impedance then add a reactive element to create the imaginary part.

Or just read the 50 ohm data into ADS and transform it there.
Like • Show 0 Likes0
Actions
xuzhu Apr 27, 2009 2:32 PM
Mark Correct
Correct Answer
Thanks all for helpful information. I really wish the network analyzer can load arbitrary port impedance into the port in the future, therefore I can do testing and tuning in real time as with 50ohm system measurement. Either loading measurement data into ADS then doing impedance conversion or doing point by point PortZ conversion is kind of slow process for my application.

Thanks so much for all you guys' help.
Like • Show 0 Likes0
Actions

Setup arbitrary frequency dependent impedance on the network

Outcomes

Related Content

Recommended Content