Contact an Expert

Thread: Help in defining calkit for TRL calibration on E5070B


Permlink Replies: 1 - Pages: 1 - Last Post: Jun 28, 2012 1:56 PM Last Post By: Dr_joel Threads: [ Previous | Next ]
mtxx

Posts: 7
Registered: 03/28/12
Help in defining calkit for TRL calibration on E5070B
Posted: Jun 24, 2012 6:46 PM
Click to report abuse...   Click to reply to this thread Reply
Previously i tried to figure out the TRL calculations so that i can measure my samples using a X-band waveguide fixture.
http://www.home.agilent.com/owc_discussions/thread.jspa?threadID=33209&tstart=0
unfortunately, its really difficult and impossible to get the job done on non agilent network analyzer.

therefore i've change my plan and made a simple parallel plate capacitor model instead.
this fixture operates at frequency much lower than X-band, therefore the E5070B is adequate for this job.

i've downloaded the vba for the TRL calibration, its suppose to allow me to define my own calkit.
i have to make my own TRL standards because i do not have a ecal or cal standards.

I've look through the user guide for this model,
http://cp.literature.agilent.com/litweb/pdf/E5070-90400.pdf

The TRL calibration is much much easier on this model, the descriptions are pretty straight forward.
define the self made calkit parameters, measure the calkit TRL and the VNA will automatically perform the calibration calculations. Amazing!

but i am stuck at one part
IMGhttp://i.imgur.com/FJk6L.png[/IMG]

i need to fill in the delays, which i do not know what are the values.

how can i measure the delays? is there any agilent guides that can help me with this?
what are the keywords i need to search?

Edited by: mtxx on Jun 24, 2012 6:46 PM

Edited by: mtxx on Jun 24, 2012 6:48 PM

Edited by: mtxx on Jun 24, 2012 6:48 PM
Dr_joel


Posts: 2,706
Registered: 12/01/05
Re: Help in defining calkit for TRL calibration on E5070B
Posted: Jun 28, 2012 1:54 PM   in response to: mtxx in response to: mtxx
Click to report abuse...   Click to reply to this thread Reply
The delays for the shorts and the thru can be zero. The delay for the line can be found by measuring the physical length of the line, in meters, and computing the time delay from delay=lenght/c=L (meters)/3e8 m/sec = delay in seconds. For X band (10 GHz) a quarter wave section is about 7.5 mm (0.0075 meters) and a delay of about 25 psec. The X band section in Agilent kits is about 32 psec, the difference being due to the no-linear phase vs frequency response of a waveguide causes the guides to be electrically longer at mid-band.; and a little error in the delay of TRL doesn't matter as the line's exact delay is computed as part of the calibration; you only need to be within about 30% of the correct answer for the most part. Do the cal, look a the delay after cal, and then reset it if it is too far off. If it is really far off, you will see some big 180 deg jumps in the phase response to clue you in.

Edited by: Dr_joel on Jun 28, 2012 12:55 PM

Point your RSS reader here for a feed of the latest messages in all forums