Hello everyone, is it possible to select the ndiff layer below the poly? [Fig 1] I draw a box over some area, then I type

select area ndiff

and to check what did I take, I move that with

move 0 -1um

and as you can see in Fig 2, only ndiff outside the poly is moved.

Magic's extraction works off of the fact that devices like FETs have their own type independent of diffusion or poly. You can do something like

select area alldiff

; however, that just moves the diffusion + transistor and leaves the poly behind. There is a roundabout way to do this, which is:

select area alldiff; select less area poly

Followed by

copy

(instead of move), then go back to the original ndiff and do

erase ndiff

. It does seem like there ought to be a simpler way to do that.

What happen if I try to remove only a contact without the diffusion below it, is it possible? Look @Tim Edwards: 1. Trying to extract only the diffusions

select area alldiff; select less area poly li; select less area ndc

then copy it [Fig 1]. The ndiffs below the contacts were not taken. 2. Removing a pdc

select area pdc; select less area pdiff
erase pdc

The pdiff below the contact is erased too. [Fig 2]

Because of the way contacts are constructed in magic, use a different command, which is

contact erase pdc

.

(2) ok. (1) The

select less area li

not only removes the contact but also the diffusion.

Sorry, I failed to notice a mistake in your command lines. . . if you want to declare multiple types at the same time, then they need to be comma-separated, with no spaces:

select area alldiff;  select less area poly,li

Review: All I wanted was the diffusions without contacts. Only with (1) -> (2) I got it, as you can see. I was hoping that with (1) -> (2) would work but it doesn't.