(a problem with hspice error) **error**: no convergence in operating point

Hi

I have a problem with my spice codes. When I add the precharge circuit, the error appears.
I have this problem only with finfet model. With other types, it works.

.option accurate=1 method=gear delmax=30p
.OPTIONS ITL1=100
.OPTIONS ITL4=100
.OPTIONS RELTOL=.01
.OPTIONS DIGSTEPBACK
.OPTIONS ABSTOL=1p VNTOL=1u
.OPTIONS METHOD=GEAR
.OPTIONS ITL5= 0
.OPTIONS ALTINIT=1

.ic V(Q)=.01
.ic V(QB)=.3

** SRAM Cell **

X1 Q QB QB 0 DGNMOS
X3 Q QB QB nvdd DGPMOS

X2 QB Q Q 0 DGNMOS
X4 QB Q Q nvdd DGPMOS
X5 bl nwl nwl Q DGNMOS
X6 blb nwl nwl QB DGNMOS


**PRECH
***********************************************
XPR nvdd npre npre BL DGPMOS
XPL nvdd npre npre BLB DGPMOS
*XPD BL 0 0 BLB DGPMOS


CBL BL 0 50F
CBLB BLB 0 50F


vvdd nvdd 0 .3

.PARAM VDD=.3

.op

vnwl nwl 0 PWL .60N 0V, 1.60N 0V, 1.62N vdd, 2.10N vdd, 2.12N 0V,

.tran .001n 6.00n

.end


Rosa

could you post the circuits?

Unfortunately you have already answered your own question.

rosaeidi said:

I have this problem only with finfet model.

Click to expand...

I don't know the finfet model, but if you only have problems with that model then it is likely to be a model problem. Can you skip the operating point calculation and go straight to rhe transient analysis? That sometimes works (although you may have to add something to the .tran statement to prevent it calculating the operating point there).

Keith.

Operating point is not important for me. I only want tranient simulation. What sholud I add to the .tran statement?

Remove the .OP and add UIC to the .TRAN so you have:

.tran .001n 6.00n uic

It means "use initial conditions" rather than calculate the operating point. It is not uncommon for simulators to use a pseudo-transient analysis to try to solve operating point convergence problems so it may work.

Keith.

When I replace .tran, the following error appears
**error**: internal timestep too small in transient analysis

Rosa

Well, the real problem is the model, but you could try the following options:

.options pivrel=0.999
.options ITL4=100
.options ITL7=1000
.options mintimestep=1p

Some of those may not be applicable to HSpice.

The best thing to do it read the section on convergence in the HSpice manual - it gives the convergence controls an their effect, and read the .OPTIONS section.

Keith.