I am looking to analyze the IL and RL of multi-board system with SAS4.0 interface.
How do I create mask in s-parameter file?
Can the SAS4 spec mask be downloaded from any Mentor forum?
The mask is not part of the S-parameter (Touchstone) file, but you can create a mask to show in the HyperLynx Touchstone Viewer. When you create masks they are saved for the specific user (and HyperLynx version). Just open the Touchstone viewer, open your Touchstone file, and create a mask from the data in your protocol specification.
Specifying the mask is pretty straightforward, but you might need to experiment with it a little to get the hang of it.
Thank you so much for the response. It is helpful.
When I see the specifications, they are generally given in the following fashion..
Is there a tool that can help us to get a tabular form?
How is mask generally created?
Awaiting your response.
Ya, Those are fun. In this situation, I use Excel to create a table of frequency - limit pairs. You only need a few values to get a good representation of the limit curve. If you find a case where your data gets very close to the sparse limit points then you can create a more detailed limit curve with more points to verify that your data still meets the limits.
W(f) is the frequency weighting function and is defined in section F.2 at page 300 of the spec.
Thank you Weston!
Appreciate your prompt response.
I tried something like that for IL. I am not sure what is [W(f)0.353] that is used for RL.
What is meant by 'W'?
I am not sure if this is the right place to ask..
I see the channel violating the mask in lower frequencies (<5.5GHz; may be because I am including cross-talk while exporting s-param?).. How is this generally fixed in PCB (without adding TAP/DFE)?
(Dotted the IL mask I created as per the excel result).
Awaiting your early response.
W(f) must be defined in the same document, probably in note 3 or 4 under the table that you showed.
The S-parameter plot that you show here is a bit unusual. The sharp resonance near 5GHz indicates a coupling to another trace, so I suggest exporting the S parameter without coupling enabled. Also, the table indicates that the limits are for differential S parameters (SDD##) so make sure you have converted the single-ended (standard) S parameters to mixed-mode before you compare the plot to the defined limits.
Thank you for the suggestion.
I will try without cross-talk and get back.
I do not have the full specification.. I could not get hold of one. This is our first time with SAS.
I'd say you really should have the full specification in order to correctly design the channel. Go to the keeper of the specification (T10 Technical Committee) and do what you need to to get the specification document.
Retrieving data ...