MOSFET Multipliers not accounted for, in SPICE Simulation

[nataraj-pinakapani]

Expected Behavior

For a MOSFET, when m=2 is specified in the SPICE Netlist, the drain current in saturation must be almost twice compared to m=1
In the given SPICE netlist I(VD2) must be twice I(VD1)

Actual Behavior

I(VD2) and I(VD1) are almost the same. This means the drain current remains the same irrespective of the value of m.

Steps to Reproduce the Problem

SPICE Netlist

** sch_path: /home/nataraj/projects/designmyic/cad/gf180mcu_invoke/xschem/sch_lib/nmos_3p3_test.sch
**.subckt nmos_3p3_test
XM1 D G GND GND nmos_3p3 L=1u W=1u nf=1 ad='int((nf+1)/2) * W/nf * 0.18u' as='int((nf+2)/2) * W/nf * 0.18u' pd='2int((nf+1)/2) * (W/nf + 0.18u)' ps='2int((nf+2)/2) * (W/nf + 0.18u)' nrd='0.18u / W' nrs='0.18u / W' sa=0 sb=0 sd=0 m=1

XM2 D2 G GND GND nmos_3p3 L=1u W=1u nf=2 ad='int((nf+1)/2) * W/nf * 0.18u' as='int((nf+2)/2) * W/nf * 0.18u' pd='2int((nf+1)/2) * (W/nf + 0.18u)' ps='2int((nf+2)/2) * (W/nf + 0.18u)' nrd='0.18u / W' nrs='0.18u / W' sa=0 sb=0 sd=0 m=2

V1 G GND 2
.save i(v1)
VD1 D GND 3.3
.save i(vd1)
VD2 D2 GND 3.3
.save i(vd2)
**** begin user architecture code

.include /home/nataraj/projects/designmyic/cad/pdk/share/pdk/gf180mcuC/libs.tech/ngspice/design.spice
.lib /home/nataraj/projects/designmyic/cad/pdk/share/pdk/gf180mcuC/libs.tech/ngspice/sm141064.spice typical

.control
save all
op
print i(vd1)
print i(vd2)
.endc

**** end user architecture code
**.ends
.GLOBAL GND
.end

Specifications

• Version: GF180MCU PDK Installed using Open PDK Version 1.0.348
• Simulator: ngspice version 37
• Platform: Ubuntu 22.04.5

[nataraj-pinakapani]

[msaligane]

Any updates on this issue?

@proppy @mithro @bmurmann

[d-m-bailey]

@nataraj-pinakapani
My understanding is that
XM1 has L=1u W=1u nf=1 m=1 which corresponds to one physical device with L=1u W=1u
XM2 has L=1u W=1u nf=2 m=2 which corresponds to 4 physical devices each with L=1u W=0.5u

Do you get the same results with
L=1u W=1u nf=1 m=2 which corresponds to 2 physical devices each with L=1u W=1u

[nataraj-pinakapani]

Hi, Sorry. I didn't work in the GF PDK for sometime now. I will check it again in the first week of August. Thanks for checking. Regards, Nataraj

…

On Wed, 19 Jul 2023 at 16:51, Mehdi Saligane ***@***.***> wrote: Any updates on this issue? @proppy <https://github.com/proppy> @mithro <https://github.com/mithro> — Reply to this email directly, view it on GitHub <#88 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AW3FYOGYNXFOCEY4PIASHATXQ77ARANCNFSM6AAAAAARMOKNYA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Cheers Natu

[ChrisZonghaoLi]

FYI this problem is now solved.

[d-m-bailey]

@ChrisZonghaoLi Great!

What was the resolution?

• a fix to ngspice?
• a change to netlist parameters?
• simulation parameters?
• fix to the mos models?

[ChrisZonghaoLi]

@ChrisZonghaoLi Great!

What was the resolution?

• a fix to ngspice?
• a change to netlist parameters?
• simulation parameters?
• fix to the mos models?

Hi Mitch, I think the original issue might be the parameter m was not declared in the spice model somehow. According to @atorkmabrains the new repo (https://github.com/mabrains/globalfoundries-pdk-libs-gf180mcu_fd_pr/tree/main/models) has the issue addressed (both ngspice and xyce models). We had this issue also discussed in Slack (https://open-source-silicon.slack.com/archives/C03R1GRP9LG/p1689280547619049?thread_ts=1689217859.856629&cid=C03R1GRP9LG).

[d-m-bailey]

Thanks @ChrisZonghaoLi.
@proppy @mithro Do you want to close this since it appears to be resolved?

[atorkmabrains]

@FaragElsayed2 Is that done?

[FaragElsayed2]

@atorkmabrains Yes

[atorkmabrains]

@FaragElsayed2 Link to PR?

[FaragElsayed2]

@atorkmabrains
mabrains/globalfoundries-pdk-libs-gf180mcu_fd_pr#152

[atorkmabrains]

@d-m-bailey @ChrisZonghaoLi Closing this ticket as it’s already addressed.