NLPLabels vs ILLabels

Ted,

If you have CDF (and are supporting ADE), then I would definitely suggest going with the ILLabels, specifically cdsName(), cdsParam(N) and cdsTerm("termName"). These have the benefit of being configurable via the annotation controls (View->Annotations in IC616, Edit->Component Display in older releases, and also via Results->Annotate in ADE). The cdsName can be used to show the instance name, cellName, both, neither; the cdsTerm can be used to display term name, net name, voltage, current, both (from DC or transient, say); the cdsParam can be used to display instance parameters, model parameters, operating point parameters, expressions of operating point parameters (in IC616) - and you can control which parameter gets displayed in which slot.

If however you want to display a fixed property that you don't want the user to have control over, you can use NLP labels too. There's no harm in having both.

Also ILLabels can be used for more complicated cases - where you need SKILL to look something up starting from ilInst.

I don't think I've used textDisplays in a symbol - not sure the UI particularly has support for this (might be used for pin names in a pin symbol - can't remember off the top of my head). So I don't see any good reason for that.

Regards,

Andrew

Thanks Andrew. I knew that there was the complicated cases requiring SKIL and ILLabels.. I was concerned about the way cdsParam specifies an index number rather than a parameter name, it is not intuitive.

For their digital library which does not use CDFs, I will use NLPLabels, their analog devices with CDFs I will use the ILLabels.

So here is a summation:

ILLabels evaluate a SKILL function or expression in the instance. The results of the function or expression is displayed as a label. They are useful if the value is a calculation.

cdsName, cdsParam and cdsTerm are functions that can be applied using ILLabels, they only seem to work in symbols, I've tried them in pcells and I'm sure that there are control properties missing. The advantage of using these functions is that the data they display is controllable in the schematic tool and can show more information than a static display such as using NLPLabels.

NLPLabels express a value found using the NLP expression language. The labels are calculated once based on information found on the instance when it is placed. A label like [@instanceName] displays the instance name using the atPar hierarchy lookup, finding the first property of that name in the hierarchy (actually on the instance). Parameter (instance property) values can be displayed but only the value is displayed and cannot be represent a calculation. The value is not displayed if it is the default value of an instance (CDF) parameter. The pin NLPLabel displays the name of the net attached to the pin. I have only tried the [@instanceName] in a pcell and it worked.

When I search the forum on NLPLabels again, after I have forgotten this, I will find it.

Cheers

Ted