Yes can you share it Pls. I tried it but I got wei...
# xschem
j
Yes can you share it Pls. I tried it but I got weird plots like this.
image.png
d
Please find attached. We observed some unexpected "kinks" for some devices for which investigations are still on-going (you will see the thread on this in the SKY slack channel)
j
Sky130 is worse in this condition, thats why I switched to gf180
d
what do you mean by "worse"? Modelling or gm/id values?
j
modelling especially svt and some gmid values
d
by "modelling" do you mean the "kink" which we have observed or have you observed some other peculiarities? What gmid values in particular? Context of my questions is I am concerned about gmid modelling in the process but only about the observed "kink". If you have seen any other strange behavior it would be great to know. Txs
j
Vt variation is very bad in svt, It was a headache to design stuff in svt
d
Interesting. Im currently characterising the vth spreads and am seeing some very large ranges also. I should be done with the full char by this week or next so will upload my results and ping you to see if they correlate with what you have seen / concluded
j
Sure lets see how bad it is
d
This is what I got. Apart from the lvt devices, spreads look pretty reasonable.
j
In one of my project I observed vth of svt nfet going as high 0.8V. I have forgotten the exact W, L
@Diarmuid Collins I tried it with gf180mcuc pfet3v3 device its very weird
d
Interesting. If you are able to re-create the svt nfet 0.8V case please let me know the dimensions. I don't do any work however with GF180 so cant comment on its pfet3v3.
j
Sure