The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> for help on convergence issue when run pss simualtion
https://designers-guide.org/forum/YaBB.pl?num=1452072065

Message started by hiSkill_11 on Jan 6^th, 2016, 1:21am

Title: for help on convergence issue when run pss simualtion
Post by hiSkill_11 on Jan 6^th, 2016, 1:21am

Hi Guys,

I am running pss+pnoise on crystal oscillator and encounter a convergence issue as following. The strang thing is that when I run the same circuit at 125deg the pss completed as expected, but fail at -40deg. I am not sure about whether it is caused by circuit or model or simulator. Any comments would be very appreciated.

*****************
DC Analysis `dcOp'
******************
Important parameter values:
reltol = 1e-06
abstol(V) = 1 uV
abstol(I) = 1 pA
abstol(Temp) = 100 uC
abstol(Pwr) = 1 nW
temp = -40 C
tnom = 25 C
tempeffects = all
gmindc = 1 pS
rabsshort = 1 mOhm
Trying `homotopy = gmin'.

Notice from spectre during DC analysis `dcOp'.
GminDC = 1 pS is large enough to noticeably affect the DC solution.
dV(I0.I1.op1) = -47.3664 uV
Use the `gmin_check' option to eliminate or expand this report.

Convergence achieved in 148 iterations.
Total time required for dc analysis `dcOp': CPU = 2.37064 s, elapsed = 937.469 ms.
Time accumulated: CPU = 4.24635 s, elapsed = 32.2662 s.
Peak resident memory used = 91.8 Mbytes.

dcOpInfo: writing operating point information to rawfile.

*****************************************************************
Periodic Steady-State Analysis `pss': estimated fund = 24.012 MHz
*****************************************************************
Trying `homotopy = gmin' for initial conditions.

Notice from spectre during IC analysis, during periodic steady state analysis `pss'.
GminDC = 1 pS is large enough to noticeably affect the DC solution.
dV(I0.I1.op1) = -47.3664 uV
Use the `gmin_check' option to eliminate or expand this report.

DC simulation time: CPU = 2.08568 s, elapsed = 542.645 ms.

MultiThread info: 7 new work threads created

Using linear IC

Warning from spectre during periodic steady state analysis `pss'.
WARNING (CMI-2139): I0.I0.N2.qsnap: The base-emitter junction current exceeds `imelt'. The results computed by Virtuoso(R) Spectre are now incorrect because the junction current model has been linearized.
WARNING (CMI-2363): I0.I0.N2.qsnap: `ic' exceeds `imax1'.
Notice from spectre during periodic steady state analysis `pss'.
I0.I0.N2.qsnap: The base-emitter junction leaves the linearized region.
I0.I0.N2.qsnap: `ic' returns to normal.
Warning from spectre during periodic steady state analysis `pss'.
WARNING (CMI-2139): I0.I0.N2.qsnap: The base-emitter junction current exceeds `imelt'. The results computed by Virtuoso(R) Spectre are now incorrect because the junction current model has been linearized.
Notice from spectre during periodic steady state analysis `pss'.
I0.I0.N2.qsnap: The base-emitter junction leaves the linearized region.

Linear IC: estimated frequency is 2.24101e+07 Hz

=================================
`pss': time = (0 s -> 2.25781 us)
=================================

Output and IC/nodeset summary:
save 4 (current)
save 20 (voltage)

Important parameter values in tstab integration:
start = 0 s
outputstart = 0 s
stop = 2.25781 us
period = 44.6228 ns
maxperiods = 50
step = 2.12394 ns
maxstep = 100 ps
ic = all
useprevic = no
skipdc = no
reltol = 1e-06
abstol(V) = 1 uV
abstol(I) = 1 pA
abstol(Temp) = 100 uC
abstol(Pwr) = 1 nW
temp = -40 C
tnom = 25 C
tempeffects = all
method = trap
lteratio = 3.5
relref = sigglobal
cmin = 1 fF
gmin = 1 pS
rabsshort = 1 mOhm

Zero diagonal found in Jacobian at `I0.I1.on' and `I0.I1.on'.
Zero diagonal found in Jacobian at `I0.I1.on' and `I0.I1.on'.
Zero diagonal found in Jacobian at `I0.I1.on' and `I0.I1.on'.
Zero diagonal found in Jacobian at `V1:p' and `V1:p'.
Zero diagonal found in Jacobian at `I0.I1.on' and `I0.I1.on'.

Error found by spectre at time = 3.2 ns during periodic steady state analysis `pss'.
ERROR (SPECTRE-16192): No convergence achieved with the minimum time step specified. Last acceptable solution computed at 0 s.

The values for those nodes that did not converge on the last Newton iteration are given below. The manner in which the convergence criteria were not satisfied is also given.
Failed test: | Value | > RelTol*Ref + AbsTol

Top 10 Solution too large Convergence failure:
I(V4:p) = 29.3536 uA, previously 29.3536 uA.
update too large: | -248.147 pA | > 29.3536 pA + 1 pA
I(V0:p) = -13.0754 uA, previously -13.0754 uA.
update too large: | -83.4838 pA | > 13.0754 pA + 1 pA
V(I0.I1.net0135) = -19.1005 V, previously -19.1005 V.
update too large: | 45.267 uV | > 19.1005 uV + 1 uV
I(V1:p) = -1.73303 mA, previously -1.73303 mA.
update too large: | 3.40771 nA | > 1.73303 nA + 1 pA
Top 10 Residue too large Convergence failure:
V(I0.I1.C9.11) = 120.642 pV, previously 120.642 pV.
residue too large: | 50 pA | > 151.123 aA + 1 pA
V(I0.I1.C0.11) = 166.275 uV, previously 166.275 uV.
residue too large: | 24.7732 pA | > 74.3217 aA + 1 pA
V(I0.I0.I0.C1.11) = -14.6418 uV, previously -14.6418 uV.
residue too large: | 12.1248 pA | > 36.3785 aA + 1 pA
V(vclk) = 3.73789 nV, previously 3.73789 nV.
residue too large: | 8 pA | > 181.01 aA + 1 pA
V(I0.I1.C19.11) = 22.1314 uV, previously 22.1314 uV.
residue too large: | 7.37346 pA | > 22.1317 aA + 1 pA
V(I0.I1.C19.1) = 731.648 mV, previously 731.648 mV.
residue too large

Title: Re: for help on convergence issue when run pss simualtion
Post by Geoffrey_Coram on Mar 16^th, 2016, 10:10am

Quote:

Zero diagonal found in Jacobian at `I0.I1.on' and `I0.I1.on'.

What's connected to I0.I1.on ? How do those devices look at -40?

Title: Re: for help on convergence issue when run pss simualtion
Post by hiSkill_11 on Mar 16^th, 2016, 11:59pm

Hi Geoffrey_Coram,

Many thanks for your help.

I attached the schematic part for you.

Geoffrey_Coram wrote on Mar 16^th, 2016, 10:10am:

Quote:

Zero diagonal found in Jacobian at `I0.I1.on' and `I0.I1.on'.

What's connected to I0.I1.on ? How do those devices look at -40?

Title: Re: for help on convergence issue when run pss simualtion
Post by Shawn Wang on Jan 7^th, 2018, 11:39pm

hi, hiSkill_11

Do you solve this problem? I encounter the same problem during PSS analysis. The tran simulation is ok, but PSS fails.

If you have the solution, could you share with me?

thanks

Title: Re: for help on convergence issue when run pss simualtion
Post by Ken Kundert on Jan 8^th, 2018, 10:17am

This appears to be a transient analysis problem rather than a PSS problem as it is occurring during the initial transient run.

I recommend that you run with the default tolerances.

-Ken