Hi, I am trying to do the layout of this circuit in magic using Tcl script. I found this script, but I don't understand which "place_nmos" / "place_pmos" correspond to. Someone can help me how identify it?

An important question is where did you "find" this Tcl script? It's not one of mine, and it doesn't belong to the PDK scripts that come with open_pdks. It is presumably intended to work within some system that has defined Tcl procedures for

place_nmos

and

place_pmos

, unless those procedures are defined earlier in your script.

This is my tcl

Thanks, that makes more sense. The script is using the cell generators in the magic PDK. It's a pretty complicated script that encodes the layout of an analog bias circuit, but is dependent on existing cells such as

cascode_bias

and

folded_cascode_p_in_stripped

which may have been created by hand or by other scripts. The

place_nmos

and

place_pmos

procedures are defined at the top of the script, and the arguments are

x_center y_center width length nf index

, where dimension values are in microns.

Okay, but I don't understand where are placed the mos in the schematic view. For example, M1 where is?

There is no obvious relationship between the schematic and script, so I don't know if there is an exact correspondence. The

index

argument to

place_nmos

or

place_pmos

is creating a device with the given index for the instance name (i.e., use index

1

for device

M1

in the schematic, which would then have the instance name

M1

in the layout. But by placing an entire cell called

cascode_bias

, it's clear that the hierarchy of the layout does not match the schematic even if the circuit does match, so there will not be a 1:1 match of device instances between the schematic and layout.

Okay, so If I want correct my script to match schematic I can use:

place_pmos x0 y0 5 1 10 16

for M16 mosfet ?