https://open-source-silicon.dev
Join Slack
Hello. I'm doing LVS to verify gf180mcuD standard cell(flip-flop and inverter) in Klayout. Howerver,...
# lvs
j
Hello. I'm doing LVS to verify gf180mcuD standard cell(flip-flop and inverter) in Klayout. Howerver, It doesn't seem to read stadard cell spice code as belows. Could you give me any advice?
m
@Junbeom Park Can you share your gds and schematic generated spice file (as opposed to just a screen shot)?
j
Thank you for your help
dffqnq_1.gdsdffqnq_1.spice
@Mitch Bailey but, .include directory should be changed in other's computer
👍 1
m
The klayout rules don’t label nwel or psub.
j
Does it means that DVNW, DVPW don't have any seperation just VDD and VSS?
m
It means that it won’t work unless there’s a tap cell to connect nwell to VDD and psub to VSS. My local klayout is not working as I wished. I’ll look into it more in the morning.
@Junbeom Park These are the problems I’m seeing so far. 1. The spice library contains 
X
devices, but klayout is extracting 
M
devices. There is a cdl library that has 
M
devices. 2. The LVS rules do not connect text to psub or nwell. 3. klayout is extracting the mosfets as 3.3V devices while the cdl defines them as 5V devices. 4. The layout is the same as dffq_1 but the source netlist has dffq_1 and an inverter. 5. There is no 
nQ
port in the layout. 6. The nfet/pfet driving net2 from the clocked inverter are slightly longer 0.605/0.505 than the schematic 0.6/0.5.
j
@Mitch Bailey Thanks for your help. 1 -> I replaced spice with cdl as below. but it brought an error in Klayout TT 3 -> how to extract mosfets as 3.3V and 5V devices both in case of simulating analog + digital part 4, 5 -> I got it 6 -> I just used standard cell provided by gf180mcuD sets. Should I modify layout of nfet/pfet driving net2?
m
1. The diodes are bad in the cdl (according to klayout). I just used cells that I needed. 3. I think there’s a recognition layer that differentiates between 3.3V and 5V. Can you find it by looking at the rule file? 6. You can change the layout or the schematic.
j
these are mos_derivations and mos_extraction file of "gf180mcu -> rule_decks"
m
@Junbeom Park What’s the difference between 
ngate_lv_n_dw
and 
ngate_5v_n_dw
?
j
@Mitch Bailey it may be "differnt layer" which divide 3.3V/5V cell as you said. I looked at the original layout of standard cell and finally I got a mistake I did when I imported cell. Contrary to the first layout picture(which I did use), the original cell has another layers like "Dualgate 55/0" and "PR_bndry0/0". I deliberately deleted these layers because they seemed not to make any difference. Thus, DRC and LVS passed. Thank you, David.
👍 1
4 Views
Open in Slack
PreviousNext
Powered by

What is Linen?

Linen is a search-engine friendly community platform. We offer integrations with existing Slack/Discord communities and make those conversations Google-searchable.

Powered by