active,p-select,n-select layer in layout

why when we draw pmos or nmos,we first draw an active layer then append a pselect or nselect layer a little larger in size to it?Why not just use p-active layer and n-active layer?
Any reason?I guess the active layer with p select and n select are just divided the one active mask to two(n and p mask) at mask manufacturing step.am I right?

anyone tell me about this？
I also found some design kit contain active,pactive,nactive,pselect,nselect simultaneously...

The answer lies in the fabrication process.
NMOS = Ndiffusion + Poly
PMOS = Pdiffusion + Poly.

Now the difference between N and P diffusion region is that of impurity (dopants ) added.

At fabrication, the initial material (.i.e ACTIVE is same for both PMOS/NMOS) while later on, this active can be converted either to be used for NMOS (ndiff ) or PMOS(pdiff). This is where the select N / P come into picture.

and it decides which diffusion to be formed.

We may use directly P-SELECT layer for the sake of design. But ACTIVE + N/P SELECT refers to fab process more closely.

HOWEVER IT VARIES FROM PROCESS TO PROCESS AND FAB TO FAB !!!

Please explain what do you mean when you say :-

"I guess the active layer with p select and n select are just divided the one active mask to two(n and p mask) at mask manufacturing step.am I right?"

I hope this helps !!

-tsb

good explanation truebs

gafsos

Some process I know only have 1 DIFF layer for you to draw. In order to distinguish N+ or P+ diffusion, we need another two layers PSELECT and NSELECT to classify.
Anyway, it's just a matter of CONVENTION, and it does not affect your fab knowledge. You will get used to it when days go

THX,your guys' comments make me a clearer picture on this but I'm still a bit confused with it....

truebs said:

"I guess the active layer with p select and n select are just divided the one active mask to two(n and p mask) at mask manufacturing step.am I right?"



-tsb

Click to expand...

I mean there is no one whole active mask during the manufactruing but actually the pselect mask and nselect mask combined as active mask?Or the fab uses one active mask and covered it with pselect mask or nselect mask to do the n or p dopant implanting?

I was really confused by this: if pselect layer or nselect layer were merely used to indicate the dopant type,why the DRC need the select layer bigger then active layer,why not just the same size?

And what about the pactive and nactive stuffs??Can it be used to instead the active-and-select style of drawing?
Also as you've mentioned,we can draw select layer to instead active layer,but the DRC need the select layer bigger than actvie layer, so r we wasting the precious silicon real estate? I don't know at this situation without active layer,is the select layer define the diffusion boundry?(I suppose when doing active-and-select drawing,the active layer define the diffusion boundry)

It's a lot of questions above,but I hope your guys can help me with my confusion...

Added after 12 minutes:

truebs said:

At fabrication, the initial material (.i.e ACTIVE is same for both PMOS/NMOS) while later on, this active can be converted either to be used for NMOS (ndiff ) or PMOS(pdiff). This is where the select N / P come into picture.

Click to expand...

Maybe the secrets lies in these words from Mrs. Truebs,but I can't grab the key...
Would u explained it in more details?
I think the active for pmos and nmos are different, and how can "active converted to either to be used for nmos and pmos"?(As it is said above)