At 09:03 PM 6/16/2012, you wrote:
I am quoting a section from the Electric user manual in Ch 7, section 7-4-2. The line in bold concerns me:




Scalable Transistors



The MOSIS CMOS technology has two transistor nodes that can take a text attribute to control their width. These transistors also have contacts built into them. Without the text attribute, the maximum width is displayed. However, by adding a "width" attribute, they shrink to that size. Note that the ports never change location, thus allowing them to scale without triggering constraints. The scaling feature of these transistors is not very useful because it is not possible to parameterize layout cells.

What does it mean that it is not possible to parametrize layout cells? From my experience, p-cells or paremetrized cells are cells where you can control some "parameter" of the device, such as length, width and such. And I see that Electric does have this capability, as I have placed MOS etc cells and indeed changed their length and width.

So could someone please explain what this statement in the manual means? Does it, maybe mean that a user cannot create a custom parametrized cells library? I hope this is not true.

Although Cadence does have "pcells" that can change their internal geometry according to parameters on the pcell, Electric does not have this capability. You cannot put a parameter onto an arbitrary Electric Cell, and then expect the geometry in that cell to alter according to the parameter value.

There are limited exceptions to this rule. For example, it *IS* possible to create parameters, especially in the schematics side of things, which will propagate down the hierarchy and be used somewhere, for example in a Spice deck. But the actual geometry, as seen on the display, cannot be affected in this way.

So the manual is telling you that the "scalable transistors" are an exception to the "no geometry can be parameterizable" rule, but a minor one, and one that doesn't give you the full power that you might want. For example, the "scalable transistors" have to be made as wide as the widest expected parameter value, because they can only really shrink from their placed size, not grow. And since other parts of the layout cannot be parameterized, the scalable transistors actually find little use in common practice.

   -Steve

--
You received this message because you are subscribed to the Google Groups "Electric 
VLSI Editor" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/electricvlsi?hl=en.

Reply via email to