I have a schematic pcell that represents an array of devices, parametrised by the number of devices in the array. Generation is fine, netlisting is good, but I'm having difficulty convincing ADE to save signals associated with the internal devices...
It's actually rather puzzling: I have two instances of this pcell, each with 4 sub-elements in the array, and I'd like to look at the currents going into the sub-elements. For instance 1, ADE seems to acknowledge that 3 of the sub-elements exist, and so lets me save those current signals.The last one cannot be saved, seemingly because ADE doesn't understand that this sub-element exists? Oddly, ADE can't find any of the 4 sub-elements associated with instance 2, so I can't save any of those signals.
A similar result occurs in the results browser after a simulation has completed. Model details for 3 out of 4 sub-elements in instance 1 and 0 out of 4 sub-elements in instance 2 exist.Is there any way to force ADE to save an output even when it can't bind the output to a net/term/instance in the design under test?
This is under 188.8.131.520.8.
I just tried selecting the outputs to be saved from the schematic, descending into the pcell, and for some reason, ADE lets me add those outputs, even though as far as I can tell, I named my signals the exact same things as what is being added.
In addition, for some of the voltage sources I'm using as current probes, I can only seem to save the MINUS port, and for others, I can save both PLUS and MINUS...
Finally, doing it manually like this only works if the current probe exists in the pcell as generated with the default parameters from the SKILL code that generates the pcell.
Thanks for any help/insight into what's going on!
In reply to kvntien:
The best way to handle this is to log the issue via customer support. Otherwise there's a lot of guesswork in understanding how your pcell behaves and trying to emulate the setup with something (hopefully) similar. If it's done via customer support, we can look at your actual data.
In reply to Andrew Beckett:
Noted, thanks for the recommendation!