causes for causality in s-parameters question

[john23]

Hello, I am using VNA with ifbw=3KHz  span 10MHz to 70GHZ step 10MHz.

I am measuring differential lines SDD11  for TDR uding IFFT.

If my IFFT TDR windows is 100ns then an event in 120ns will go to -20ns thus will create causality issue.

Anothe cause I know is deembeding creating ringing ,but i am not doing deembeding.

But  this is all good in theory. In the lab what could cause causality issues in the s-params? how can I handle them?

Thanks. 

[ejeffrey]

Obviously nothing in the "lab" can cause acausal response.  Well, unless you drove your delorean to the lab this morning.

Whatever is happening to make it look like that is from signal processing.  Its fairly easy amd common.  For instance if you apply a bandwidth limiting filter and then calibrate so that the response peak is zero delay you will get substantial leading response.  Thats no problem as long as you understand what it means. 

[ejeffrey]

,but i am not doing deembeding.

Deembedding is like using timezone.  You always are, it's just a question of whether make the choice consciously or have the choice made for you.

[john23]

Hello ejeffrey,this is very intresting.I am talking purely on measuring PCB SDD11 for TDR  there is no extra filter component.
purely speaking in VNA terms.
what could do BW limitation?
could you give more examples?

Where I replace the cables connected to the VNA I see better causality and I dont know why I calibarated evey cable set when I connect it to the VNA.

[booscrawl]

The VNA uses repeated pulses to measure things. If something rings for a time duration similar to the period between pulses, it will look like the response happens before the pulse