calibration discontinuity at the frequency break point,

VNA calibration

Hi, I'm a student of Electronic engineering in Rome, Università Tor Vergata. I'm writing my thesis on "On wafer calibration techniques" and I'm working in the laboratory with the HP8510C Vector Network Anlyzer. I need, if possible, an answer to these questions:

1) Does HP8510C's firmware include LRRM algorithm ?

2) Is HP8510C able to distinguish, during the calibration sequence, (and how does it) the match standard used for low frequencies from the line standards in the TRL calibration procedure ?

3) I have studied the most common on wafer calibration techniques: SOLT, TRL, LRM, LRRM. What other non coaxial calibration procedures should I study ?

4) Is it possible to use LNN, LRR, MULTILINE for non coaxial calibrations ?

Thank you very much for your help.

VNA calibration

1) NO
2) what do you mean? Are you speaking about the use of a Line infinitely long (so non reflecting)?
3)...This answer is didactic rather than technical.
4) using 8510 is possible donwload and upload the error coefficients. Many off-line applied calibration algoritms are applicable, but the question is: there is a really improvement or a step down on the uncertainty?

Re: VNA calibration

The multiline TRL calibration , developed by NIST researchers, seems to be the most accurate calibration (less uncertainty!) .

Re: VNA calibration

on question 2)=>

the network analyzer doesn't know anything about your standard, it knows (measure) the incident/reflect signal only. So if you connect to "short" when it asks for a "match", it will take it as a match. So definition of "standard" is your job: it has to be consistent with your de-embedding algorithm and the "thing" that you want to measure.

It is useless make such claim as to which method is the best without specifying the test platform. In some case TRL is simply impossible.

Re: VNA calibration

francut78 said:

]
......I'm writing my thesis on "On wafer calibration techniques" and ....

Click to expand...

loucy said:

on question 2)=>
....It is useless make such claim as to which method is the best without specifying the test platform. In some case TRL is simply impossible.

Click to expand...

TRL cal was born for on wafer measurements.