How to preserve parasitic nodes in extracted spice netlist with APS++/parasitic reduction

TJ,

It's more important to know which spectre version you're using (the IC version is irrelevant in this case). You'll see that at the top of the spectre.out log file.

From some quick searching, this should work from an MMSIM14.1 ISR (not sure which version), but I see there were some fixes added in MMSIM15.1 (that was actually related to the save statements turning off reduction altogether in some cases which it shouldn't have done). 

The advice is to ensure that you put sigtype=node on the save statement with the wildcard, but I'd first make sure you're using a recent enough MMSIM version.

Kind Regards,

Andrew.

Hi, Andrew,

I put a very tiny donut like metal resistor with an schematic instance name "probe" on a metal line in the layout so the extract will put this parasitic resistor in the extracted netlist with identifiable net and instance names and I can save the probe point by "save *probe* sigtype=node". I used the instance preservation option to preserve this resistor and it is trimmed off. Please see the log and part of the extracted netlist for the detail. One detail is that one terminal of this resistor taps on the metal line that I am interested in and the other terminal is dangling and connected to nowhere.

This is part of Spectre log:

Cadence (R) Virtuoso (R) Spectre (R) Circuit Simulator Version 15.1.0.345.isr2 64bit -- 4 Jan 2016

Command line:
/opt/beta/tools/artist/MMSIM_15.10.345/tools.lnx86/bin/spectre -64 \
input.scs +escchars +log ../psf/spectre.out -format psfxl -raw \
../psf +aps +parasitics +lqtimeout 900 -maxw 500 -maxn 500 \
-maxwtl 500 -maxntl 500 +lsuspend +lqt 0 +lqs 30

Global user options:
reltol = 0.001
vabstol = 1e-06
iabstol = 1e-12
temp = 27
tnom = 25
scalem = 1
scale = 1
gmin = 1e-12
rforce = 1
maxnotes = 500
maxwarns = 500
maxwarnstologfile = 500
maxnotestologfile = 500
digits = 5
cols = 80
pivrel = 0.001
sensfile = ../psf/sens.output
checklimitdest = psf
preserve_inst = *Rprobe
save = selected
subcktprobelvl = 2

Scoped user options:

Circuit inventory:
nodes 1448
capacitor 4178
juncap200 2
resistor 4661
utsoi2 40
vsource 3

Analysis and control statement inventory:
info 7
tran 1

Output statements:
.probe 0
.measure 0
save 2

Notice from spectre during initial setup.
APS Enabled.
Parasitics Reduction Enabled.
(Resistors reduced by 89.07% Capacitors reduced by 65.48%)

Wildcard match summary:
save *probe* nodes: 0
save *Rprobe nodes: 0

Output and IC/nodeset summary:
save 1 (voltage)

Part of input.scs

*|NET XI0/XINP_probe/vprobe 2.72049e-16

**|OP (XI0/XINP_probe/Rprobe:pos XI0/XINP_probe/Rprobe pos B 0 1.88989 2.174)

C4_1403 XI0/MN9@2:G XI0/XINP_probe/Rprobe:pos 6.99675e-18

R4_1775 XI0/XINP_probe/Rprobe:neg INP:1 0.292624

RXI0/XINP_probe/Rprobe XI0/XINP_probe/Rprobe:pos XI0/XINP_probe/Rprobe:neg r=0.001

simulatorOptions options reltol=1e-3 vabstol=1e-6 iabstol=1e-12 temp=27 \

   tnom=25 scalem=1.0 scale=1.0 gmin=1e-12 rforce=1 maxnotes=500 \

   maxwarns=500 maxwarnstologfile=500 maxnotestologfile=500 digits=5 \

   cols=80 pivrel=1e-3 sensfile="../psf/sens.output" checklimitdest=psf \

   preserve_inst=[*Rprobe]

tran tran stop=3n write="spectre.ic" writefinal="spectre.fc" \

   annotate=status maxiters=5

finalTimeOP info what=oppoint where=rawfile

modelParameter info what=models where=rawfile

element info what=inst where=rawfile

outputParameter info what=output where=rawfile

designParamVals info what=parameters where=rawfile

primitives info what=primitives where=rawfile

subckts info what=subckts where=rawfile

save net1

saveOptions options save=selected subcktprobelvl=2

save *probe* sigtype=node

save  *Rprobe:currents

You might be being stymied by the dangling resistor removal in APS. Not sure. Given that you're using a recent version, I'd suggest contacting customer support so that we can explore this in more detail to find out why it's not working or if there's a workaround (maybe you can share the example that has the problem which would make it quicker to reproduce).

Thanks,

Andrew.

That article (solution) isn't really relevant to this - but anyway, hopefully you'll get some response via this case soon enough. There's an AE assigned to work on it.

Regards,

Andrew.