Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Problem with MOS Bulk Routing Parasitic Resistance Extraction

Status
Not open for further replies.

Stepan Sutula

Newbie level 3
Joined
Feb 2, 2015
Messages
3
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
31
Hi,

During the Spectre netlist generation process, I have found out that the Assura QRC extractor includes the parasitic resistances of the MOS transistor bulk routing in the extraction view:
rg58 (vdda \5\:vdda) resistor r=24.5376 c=0
But, it skips connecting this parasitic resistor to the bulk of the transistor:
M10 (\3\:vout \5\:nclk \2\:vin vdda) p_lv_18_mm w=1.92e-06 l=2.4e-07 ad=2.224e-13 as=2.224e-13 pd=2.0u ps=2.0u m=(1)*(1)
From the line above, you can see that it circumvents the parasitic resistor and connects the bulk directly to the polarization net vdda (NOT \5\:vdda)
As a result, the simulation netlist generated from this extracted view is also incorrect.

I have tried two different technologies with two different design kits and both have the same issue. Two kits are:
1. Virtuoso Design Environment version IC6.1.3.500.13 with Cadence Extraction QRC - Parasitic Extractor - Version 8.1.4-p002.
2. Virtuoso Design Environment version IC6.1.5-64b.500.132 with Cadence Extraction QRC - Parasitic Extractor - Version 11.1.2-p106.

Is there any way to fix this problem? Can it be done by modifying the configuration script or updating the kit?

Thanks,
Stepan.
 

There are more than one ways to do that.
QRC substrate handling methodology has been changed sin 8.x [exact version number I forgot] - there are detailed and good technical reasons backed by many big design house and fab. In new versions of QRC, the back-gate of MOS will get ideal VSS/VDD [based on psub,nwell in NMOS/PMOS] - by default, unless proper methodology is in place.

Few recommendations :
- Check your PDK docs [Readme etc]
- Check LVS extract rule [if substrate is stamped - check next line - else skip next line]
- Look for stamp=1 or stamp=2 stuffs in p2lvsfile /or/ layer_setup in QRC-tech-dir [ if not there, you can add and do Techgen compilation process once again ]
- If above is not there - use ignorevias thing.. [ this has to do with how substrate gets the subnet - even when not stamped ]

In physics, substrate parasitics separately can also be extracted - it is a 3D mesh - which needs foundry provided doping profiles and so on.. QRC is the only product that has that capability at the chip level - been used by some RF design houses.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top