AnsweredAssumed Answered

In-Fixture TRL Calibration

Question asked by mirza on Feb 19, 2013
Latest reply on Feb 27, 2013 by Dr_joel
Hello! I would really appreciate any help. I read the Agilent application notes, but still have some doubts about the TRL calibration process.

We just purchased a commercial RF small-signal test fixture with in-fixture TRL calibration. They have provided us with the following information regarding the calibration standards.

1. LINE
start freq: 2 GHz
stop freq: 18 GHz
delay: 25 ps
line length: 67.90 mm



2. THRU
length: 63.5 mm
reference length: 31.75 mm
(so I assume it is a zero length THRU)

3. REFLECT
a short with 0 ps delay


The insertion phases are 18 degrees and 162 degrees. Since they are very close to 20 and 160, I assume they are alright).

The network analyzer I am using is Agilent PNA E8364B. My questions are:

1. Is this information enough for the calibration? Because, for example, when asked for L0, L1, L2, L3 in the definition for short; I just put 0.
2. I also put 0 in the field for loss for all three standards.
3. Can I use this calibration kit for frequencies below 2 GHz? For Example: 50 MHz to 6 GHz.
4. I read that the loss for LINE can be approximated by gating. But I am not sure how to do it. Do I put the LINE after calibration and then see the time domain response?

I performed a calibration from 2 - 6 GHz and later measured the S21 response for the THRU standard. I got these following values:

Mag: around 1.3 dB
Phase: around -5 Degree

So I am not very happy with the quality of the calibration since both values should be close to zero. Is it because of the lack of any information in the calibration standards or is it because of my expertise?

I would really appreciate any help regarding this. Thanks in advance.

Edited by: Mirza on Feb 19, 2013 1:21 PM  

Outcomes