ECAL issue: Number of ports must agree

[Michael Smith]

Hello!

I'm running into a roadblock performing error correction on the various impedance states of my Keysight Ecal module. The error I'm receiving is "number of ports must agree." The console doesn't tell me much information to try debugging. Here are my steps:

1. Perform a 2-port cal using mechanical kit

2. Use Measurement series to capture the impedance states of the ECU (A1-A7, B1-B7, AB1, AB2) by connecting port 1 to Ecal Port A and port 2 to Ecal Port B.

3. I then save these error corrected impedance states as a databased calibration standard - one file for each state. I have saved them with the default ports (1, 2), with only Port A states for port 1 and Port B states as port 2, and as empty (all ports).

4. I then perform a "calibration"  using the Ecal module by performing an ECU  measurement series and then saving those states under a new project/journal as cal standards.

5. In the cal config, I define a new SOLT cal and then load all the reflection terms for A1-A7 for port 1, all the reflection terms for B1-B7 for port 2, and then transmission 1-2 for AB1 and AB2. I have also tried an Optimized Cal.

6. I save the configuration and then start the computation.

Each time VNA tools (latest version 2.5.3) runs thru the computation, I get an error that the number of ports must agree. I have tried everything I can think of, followed the instructions on this forum, and also from what I learned in the VNA Tools workshop but I'm still running into issues. I'm sure there is an obvious and easy solution and I sure would appreciate hearing it!

Thank you!

Michael Smith

[Michael Wollensack METAS]

Hi,

the procedure looks ok for me. 

For the one port Ecal states A1-A7 you should only store port 1 of your two port data in the database, as you did as well.

For the one port Ecal states B1-B7 you should only store port 2 of your two port data in the database, as you did as well.

Can you please post the error which you see in the console window including the stack trace. This helps me do see where it fails.

Regards

Michael

[Michael Smith]

Thanks Michael. I'm going to post that up here in just about two weeks. I just left for a couple weeks vacation.

Michael

[Michael Smith]

Good news, Michael. I was able to save the impedance states as databased standards and refer to them on a new calibration. I learned that AB1 and AB2 need to be associated with their corresponding databased cal standard entry as opposed to using ideal 50 ohm thru. Otherwise, you get a significant DC shift in the error corrected response. However, I am still have some issues which I will open a new subject to address.

Michael