Does TRL calibration need to define offset loss of thru and line?

[macgyfu]

Hi All
I make a set of coaxial cable TRL calibration kit. but I have a little confuse of defining the parameter of calibration kit.
I use Agilent E5071C network analyzer. The below example was from Agilent website. It tells me I need to define the offset loss of thru and line. But in my understanding, TRL theory should not need any known parameter of calibration kit, even the delay of line.
It is a little weird, If it is necessary, how do I define it? there are two different coaxial cable from NA to DUT, one is "expensive cable" which it was used to being reference plan of SOLT calibration. And the other one is TRL calibration kit cable. so I have two different cables if I need to measure offset loss. Do I measure cable1, cable2 or cable1+cable2?
"
Example of defining the TRL calibration kit
You need to enter the definition of the TRL calibration kit to perform TRL calibration. Follow these steps to define the following calibration kit given as an example.
THRU (Delay 0 ps, Offset Loss 1.3 Gohm/s)
REFLECT (SHORT, Delay 0 ps)
MATCH (@0 to 2GHz)
LINE1 (50-ohm transmission line, Delay 54.0 ps @ 2G to 7GHz)
LINE2 (50-ohm transmission line, Delay 13.0 ps @ 7G to 32GHz)
"
Please somebody help me to release my confuse
Have a nice day
MacGyFu

 

[vfone]

I don't understand how you want to do a TDR measurement using the VNA, but I am not interested to find.
By definition TDR measurements use an oscilloscope and a pulse generator. Check the application note below from Tektronix, which is one of the inventors of this method.
**broken link removed**

 

[macgyfu]

I mean TRL calibration, not TDR measurement.

I don't understand how you want to do a TDR measurement using the VNA, but I am not interested to find.
By definition TDR measurements use an oscilloscope and a pulse generator. Check the application note below from Tektronix, which is one of the inventors of this method.
**broken link removed**

 

[vfone]

Sorry, my mistake...

 

[Mazz]

MacGyFu

I don't understand exactly what is your issue.

Some remarks: consider the VNA+the expensive cable as a single unit. Finishing TRL cal, your reference plane will be at this single unit (whatever it is, with a single expensive cable and VNA connector, or any other cable (not TRL cal kit) you are using to connect your DUT).

In TRL the electrical length need only be specified within 1/4 wavelength, that's why the VNA wants to know the electrical length.

Do some tests of calibration using different delay definitions and check them, to appreciate the differences, if any.

Mazz

 

[eleazar_777]

HI MacGyFu

The TRL calibration technique need the Line impedance and the length of the lines, I can attach you the LIMCAL software, this is a multi line TRL calibration program may be can help you.

It is very easy to use you need to follow the next steps

1.- Calibrate the VNA with SOLT as a first tier.
2.- Measure your TRL standar and your dut (second tier) and put that files (*.s2p) in to the program
3.- Calibrate the dut files and thats it

 

[Mazz]

eleazar_777

this is interesting; is it possible to have this software?

Mazz

 

[macgyfu]

Hi Mazz
I can understand why the LINE must be specified within specific frequency range (some document says the opeartion range is between 20 t0 160 degrees) to get more accurate calibration result. When we are doing TRL calibration, we will tell NA the range of each LINE, for example LINE1 2GHz to 4GHz. That's the point I don't understand why we need to give the delay length of LINE to NA, it seems a little redundant.
But I don't understand why OPEN is limited within 1/4 wavelength as well.
I quoted a sentence below from Agilent as attached link.
It says Phase of G must be know within +-1/4 wavelength. When we do TRL calibration, The reference plane is a coxial cable + TRL_kit OPEN length. The total physical length of the coxial cable + TRL_kit open length is very long, it means the f0 of 1/4 wavelength is very low. It is unreasonable there is such limitation in TRL measurement, so I guess I misunderstand something.
Does anyone know the meaning of this sentence?:|
MacGyFu
"
REFLECT:
Reflction coefficient G magnitude (optimally 1.0) need not be known
Phase of G must be know within +-1/4 wavelength *1.
Must be the smae G on both ports
May be used to set the reference plane if the phase response of the REFLECT is well-know and specified.
*1: The phase response need only be specified within a 1/4 wavelength +-90 degrees either way. During computation of the error model, the root choice in the solution of a quadratic equation is made based on the reflection data.
An error in definition would show up as a 180-degree error in the measured phase.
"
http://emlab.uiuc.edu/ece451/appnotes/TRL_2.pdf

MacGyFu

I don't understand exactly what is your issue.

Some remarks: consider the VNA+the expensive cable as a single unit. Finishing TRL cal, your reference plane will be at this single unit (whatever it is, with a single expensive cable and VNA connector, or any other cable (not TRL cal kit) you are using to connect your DUT).

In TRL the electrical length need only be specified within 1/4 wavelength, that's why the VNA wants to know the electrical length.

Do some tests of calibration using different delay definitions and check them, to appreciate the differences, if any.

Mazz

 

[Mazz]

macgyfu

first, it seems strange that in their cal kit they use an OPEN as REFLECT standard. Usually a SHORT is better, but maybe in coax (I work with planar lines) this is not true.

Anyway, the reason for this sentence is, as far as I can understand, that the reference plane for S11 phase has to be determined by the REFLECT, so you have to tell to the NA its lenght. 1/4 wavelenght is obvious, this will avoid the NA to confuse an OPEN with a SHORT or viceversa.

Do the test I suggested with you cal kit: use a wrong definition of LINE delay a see the difference it makes in cal verification. You will appreciate the importance of this parameter.

Mazz