Does anyone know about deep nwell in gf180mcu? Is ...
# gf180mcu
m
Does anyone know about deep nwell in gf180mcu? Is the nwell guard ring calculated automatically on mask creation? If not, what are the overlap rules? I couldn’t find any documentation.
r
Nwell min. enclosure by Dnwell 0.5um. Don’t see anything else.
a
You need to draw 2 layers: • DNWELL • LVPWELL
Large DNWELL enclosing smaller LVPWELL This will create the isolated LVPWELL for you.
@Mitch Bailey I hope this answer your question. Let me know if you have more questions.
m
Thanks @Amro Tork so for gf180mcu, there is no need for nwell guard ring on dnwell? The nwell ring is auto generated? How does this work with mask derivations? PWELL = not (NWELL or DNWELL ) or LVPWELL? How does the manufacturing process work psubstrate pwell? deep nwell nwell lvpwell Do you have the gds -> mask conversion formulas?
a
Yes, I believe NWELL are both partially derived and some drawn for GF180MCU
There is another document that is not part of the Open Source release that describes the MASK generation process
Check the link above
I hope that makes sense
We always assume that we start with p-sub
Wafers need a couple of steps to get the p-sub: • Initial epitaxial layer growth. • Initial Pwell deposition
m
I saw the mask entries in the pdk and noticed the drawn and generated. I want to know the formula for generation.
a
It’s in a different document
And I believe they won’t share it publicly as it has mask preparation information
m
Not one source, right?
a
What do you mean by not one source?
m
*open
a
Yes
t
@Mitch Bailey: The GF process deep nwell is fundamentally different from the sky130 deep nwell. There is no nwell "boundary layer" around the deep nwell. The deep nwell is considered by default to be n-type all the way up to the surface. In sky130, a region inside the nwell boundary ring is automatically p-type. In GF, any region inside the deep nwell is automatically n-type unless a pwell layer is drawn.
m
Thanks, Tim. Now I need to track down everybody I told differently. 🤢