When using mtline model in cadence spectre schematic simulation, why I
can not connect output of one mtline to output of another mtline?
(warnings show there will be short circuit and floating nodes)
This vaguely rings a bell.
Where is the warning appearing, and what exactly is the message you see?
Which IC and MMSIM version are you using? (the spectre version will appear in the spectre log file at the top, and the IC version can be found by using Help->About in the CIW, or by typing getVersion(t) in the CIW).
In reply to Andrew Beckett:
In reply to jyang:
Please show a picture of the schematic (use the Option tab on the forum to upload it) and the precise messages from the CIW (please cut and paste them).
The problem is something other than what I was remembering (which was fixed a long time ago)
warning message in CIW
It's because the symbol has the pins on the left as being input pins, and those on the right as being output pins.
It won't affect the simulation - it's just a spurious check because the pin directions are set on mtline. It would be better if mtline had inputOutput pins as each pin - then the problem wouldn't occur. So I'd suggest raising this with customer support.
You can go to Check->Rules and set "Shorted Output Pins" to "ignored" and then the warnings will go away.
This has been submitted as CCR 1156640 (Title: Pins of mtline should have inputOutput direction). Anyone who also has a problem with the pin directions of the mtline component should consider asking Cadence support to file a duplicate of this CCR on his behalf. If more people are asking for this change, it is more likely to be implemented soon.
In reply to Frank Wiedmann:
I have just received the following email from Cadence:
Hello Frank,Regarding Case 45446581, "Change pins of mtline in analogLib to inputOutput." I wanted to inform you that your Cadence Change Request (CCR) number status has been changed.The new status is set to Inactive, which means that no action is planned. EachCCR is carefully considered, evaluated, and prioritized along with other fixes,planned feature additions, and enhancement requests, for possible inclusion inupcoming product updates and releases.If you feel this is very important and would like us to reconsider, respond to thismessage and the AE working on your Case will be in contact with you.Regards,Cadence Customer Support
I am a bit surprised about this as I would have assumed that this change could be made in a few minutes. Anyone who still would like to have this changed should ask Cadence to file a duplicate of CCR 1156640 on his behalf.
First of all, there really is no such thing as a "few minute" fix. Fixes need to be implemented, reviewed, checked in and potentially merged into multiple streams, tested, regression tests implemented - all of which takes more time. Given that we end up with lots of small fixes such as this which are competing for attention, we tend to focus on those which are the biggest obstacles to be fixed in hotfixes. Other things will be planned for future releases during the planning phase for a release. If submitted before that planning phase, or even if the plate is too full for a particular release, they may be marked Inactive-Backlog. These are then reconsidered at the next planning phase - we cannot plan the next release as we go along as otherwise it would always be first-in first-out rather than assessing the overall requirements.
So no need to campaign for this in public; it's not being ignored.
Thanks a lot for the clarification. Unfortunately, the wording of that email gave me the impression that Cadence did not intend to ever implement this change at all.