steveb@hobiecat.cs.caltech.edu (Steve Burns) (11/01/90)
I've had some difficulty simulating asynchronous circuits in SPICE3C1 under the level4 (BSIM) model. The message returned is: doAnalyses: Timestep too small I am familiar with messages of this type having used SPICE2G6. However, in SPICE3C1, under the level2 model, I no longer have this convergence problem. Is it possible that the non-linear capacitance fixes incorporated into SPICE3 were only added to the level1, level2 and level3 MOS models, but not to the level4 MOS model? If this is the case, does anyone have patches to the level4 model that fix this problem? Two SPICE decks follow that illustrate this problem. The circuits are identical (essentially ring oscillators) and the model parameters come from the 2 micron MOSIS run M8CH. ============= level4 --- convergence problems =========================== C ELEMENT --- TWO INVERTERS --- BSIM .model p pmos level=4 + vfb = -.24212 lvfb = -0.014567 wvfb = .333824 + phi = .659278 lphi = 0 wphi = 0 + k1 = .595438 lk1 = -0.043545 wk1 = -.37067 + k2 = 0.0125327 lk2 = 0.065523 wk2 = -.1688 + eta = -0.0050896 leta = 0.0493218 weta = -0.011635 + muz = 229.792 dl = .37685 dw = -.27143 + u0 = .126809 lu0 = 0.0470339 wu0 = -.1009 + u1 = 0.00414901 lu1 = .317906 wu1 = -0.035473 + x2mz = 9.90559 lx2mz = -4.6505 wx2mz = 3.87995 + x2e = -0.00035035 lx2e = -0.0031942 wx2e = -0.0047378 + x3e = -0.00098068 lx3e = 0.00115458 wx3e = -0.0046737 + x2u0 = 0.00587181 lx2u0 = -0.0022628 wx2u0 = 0.000532583 + x2u1 = -0.00049022 lx2u1 = 0.0050853 wx2u1 = 0.00865839 + mus = 233.995 lmus = 170.218 wmus = -44.688 + x2ms = 8.44038 lx2ms = 2.65093 wx2ms = 6.12955 + x3ms = 0.01877 lx3ms = 11.3848 wx3ms = 2.67881 + x3u1 = -0.014441 lx3u1 = -0.0055355 wx3u1 = 0.0181036 + tox = 0.0403 temp = 27 vdd = 5 + cgdo = 1.61447e-10 cgso = 1.61447e-10 cgbo = -2.32568e-10 + xpart = 1 + n0 = 1 ln0 = 0 wn0 = 0 + nb = 0 lnb = 0 wnb = 0 + nd = 0 lnd = 0 wnd = 0 + rsh = 115.8 cj = 0.0002846 cjsw = 2.95e-10 + js = 1e-08 pb = .53 pbsw = .18 + mj = .5126 mjsw = .1915 wdf = 0 + dell = 0 .model n nmos level=4 + vfb = -.80414 lvfb = -.24896 wvfb = .232322 + phi = .751816 lphi = 0 wphi = 0 + k1 = 1.07071 lk1 = -0.029794 wk1 = .378009 + k2 = .123487 lk2 = .153831 wk2 = -0.067804 + eta = -0.0047538 leta = 0.0386962 weta = -0.0078346 + muz = 569.029 dl = .251073 dw = -.81408 + u0 = 0.038669 lu0 = 0.0479726 wu0 = -0.037994 + u1 = -0.0038072 lu1 = 1.22724 wu1 = -.49543 + x2mz = 7.47361 lx2mz = -10.386 wx2mz = 55.928 + x2e = -0.00033008 lx2e = -0.016979 wx2e = 0.00339123 + x3e = -2.1618e-05 lx3e = -0.0046946 wx3e = -0.0032476 + x2u0 = 0.00164666 lx2u0 = -0.0030019 wx2u0 = 0.021817 + x2u1 = -0.02083 lx2u1 = 0.0433833 wx2u1 = 0.0644895 + mus = 589.982 lmus = 816.976 wmus = -435.41 + x2ms = -10.39 lx2ms = 43.9484 wx2ms = 113.076 + x3ms = -.53774 lx3ms = 118.454 wx3ms = -63.724 + x3u1 = 0.0011957 lx3u1 = .12558 wx3u1 = -0.060857 + tox = 0.0403 temp = 27 vdd = 5 + cgdo = 1.07563e-10 cgso = 1.07563e-10 cgbo = -6.97523e-10 + xpart = 1 + n0 = 1 ln0 = 0 wn0 = 0 + nb = 0 lnb = 0 wnb = 0 + nd = 0 lnd = 0 wnd = 0 + rsh = 37.22 cj = 0.0001666 cjsw = 5.075e-10 + js = 1e-08 pb = .35 pbsw = .8 + mj = .5838 mjsw = .2987 wdf = 0 + dell = 0 * * "x" 2 c2 2 0 9.72fF * "y" 3 c3 3 0 9.72fF * "z" 4 c4 4 0 9.72fF * "reset" 5 c5 5 0 9.72fF * "w" 6 c6 6 0 9.72fF * M0 7 2 0 0 n L=2.0u W=25.2u PD=2.0u PS=2.0u AD=25.2p AS=25.2p +NRD=0.0397 NRS=0.0397 M1 4 3 7 0 n L=2.0u W=13.6u PD=23.6u PS=2.0u AD=68.2p AS=13.6p +NRD=0.3666 NRS=0.0733 M2 8 5 1 1 p L=2.0u W=40.0u PD=2.0u PS=2.0u AD=40.0p AS=40.0p +NRD=0.0250 NRS=0.0250 M3 6 4 8 1 p L=2.0u W=23.1u PD=33.1u PS=2.0u AD=115.7p AS=23.1p +NRD=0.2160 NRS=0.0432 M4 2 6 0 0 n L=2.0u W=16.9u PD=26.9u PS=2.0u AD=84.4p AS=16.9p +NRD=0.2961 NRS=0.0592 M5 3 6 0 0 n L=2.0u W=6.8u PD=16.8u PS=2.0u AD=34.0p AS=6.8p +NRD=0.7362 NRS=0.1472 M6 9 2 1 1 p L=2.0u W=40.6u PD=2.0u PS=2.0u AD=40.6p AS=40.6p +NRD=0.0246 NRS=0.0246 M7 4 3 9 1 p L=2.0u W=21.3u PD=31.3u PS=2.0u AD=106.6p AS=21.3p +NRD=0.2346 NRS=0.0469 M8 6 4 0 0 n L=2.0u W=14.6u PD=24.6u PS=2.0u AD=73.2p AS=14.6p +NRD=0.3416 NRS=0.0683 M9 6 5 0 0 n L=2.0u W=4.0u PD=14.0u PS=2.0u AD=20.0p AS=4.0p +NRD=1.2500 NRS=0.2500 M10 2 6 1 1 p L=2.0u W=25.4u PD=35.4u PS=2.0u AD=127.2p AS=25.4p +NRD=0.1965 NRS=0.0393 M11 3 6 1 1 p L=2.0u W=12.3u PD=22.3u PS=2.0u AD=61.6p AS=12.3p +NRD=0.4061 NRS=0.0812 vdd 1 0 dc 5v vreset 5 0 pulse ( 5v 0v 2ns 0.2ns 0.2ns 1000ns 1000ns ) .tran 0.1ns 20ns .plot tran v(2) v(3) v(4) v(5) v(6) (-.2,5.2) .end ========================================================================= ============= level2 --- no convergence problems ======================== C ELEMENT --- TWO INVERTERS --- LEVEL2 .MODEL n NMOS LEVEL=2 LD=0.292923U TOX=403.000026E-10 + NSUB=1.252361E+16 VTO=0.738 KP=5.544000E-05 GAMMA=0.7551 + PHI=0.6 UO=646.871 UEXP=0.195772 UCRIT=23846.8 + DELTA=1.56163 VMAX=53213 XJ=0.250000U LAMBDA=4.249627E-02 + NFS=1.083036E+12 NEFF=1 NSS=1.000000E+12 TPG=1.000000 + RSH=37.220003 CGDO=2.509834E-10 CGSO=2.509834E-10 CGBO=-5.703876E-10 + CJ=1.666000E-04 MJ=0.583800 CJSW=5.513000E-10 MJSW=0.247000 PB=0.350000 * Weff = Wdrawn - Delta_W * The suggested Delta_W is -0.67 um .MODEL P PMOS LEVEL=2 LD=0.206151U TOX=403.000026E-10 + NSUB=8.924787E+15 VTO=-0.885081 KP=2.408000E-05 GAMMA=0.6375 + PHI=0.6 UO=281 UEXP=0.260883 UCRIT=18961.6 + DELTA=0.299559 VMAX=47630.2 XJ=0.250000U LAMBDA=5.789001E-02 + NFS=5.583092E+11 NEFF=1.001 NSS=1.000000E+12 TPG=-1.000000 + RSH=115.800006 CGDO=1.766351E-10 CGSO=1.766351E-10 CGBO=1.406048E-10 + CJ=2.846000E-04 MJ=0.512600 CJSW=2.639000E-10 MJSW=0.227200 PB=0.530000 * Weff = Wdrawn - Delta_W * The suggested Delta_W is 0.16 um * * "x" 2 c2 2 0 9.72fF * "y" 3 c3 3 0 9.72fF * "z" 4 c4 4 0 9.72fF * "reset" 5 c5 5 0 9.72fF * "w" 6 c6 6 0 9.72fF * M0 7 2 0 0 n L=2.0u W=25.2u PD=2.0u PS=2.0u AD=25.2p AS=25.2p +NRD=0.0397 NRS=0.0397 M1 4 3 7 0 n L=2.0u W=13.6u PD=23.6u PS=2.0u AD=68.2p AS=13.6p +NRD=0.3666 NRS=0.0733 M2 8 5 1 1 p L=2.0u W=40.0u PD=2.0u PS=2.0u AD=40.0p AS=40.0p +NRD=0.0250 NRS=0.0250 M3 6 4 8 1 p L=2.0u W=23.1u PD=33.1u PS=2.0u AD=115.7p AS=23.1p +NRD=0.2160 NRS=0.0432 M4 2 6 0 0 n L=2.0u W=16.9u PD=26.9u PS=2.0u AD=84.4p AS=16.9p +NRD=0.2961 NRS=0.0592 M5 3 6 0 0 n L=2.0u W=6.8u PD=16.8u PS=2.0u AD=34.0p AS=6.8p +NRD=0.7362 NRS=0.1472 M6 9 2 1 1 p L=2.0u W=40.6u PD=2.0u PS=2.0u AD=40.6p AS=40.6p +NRD=0.0246 NRS=0.0246 M7 4 3 9 1 p L=2.0u W=21.3u PD=31.3u PS=2.0u AD=106.6p AS=21.3p +NRD=0.2346 NRS=0.0469 M8 6 4 0 0 n L=2.0u W=14.6u PD=24.6u PS=2.0u AD=73.2p AS=14.6p +NRD=0.3416 NRS=0.0683 M9 6 5 0 0 n L=2.0u W=4.0u PD=14.0u PS=2.0u AD=20.0p AS=4.0p +NRD=1.2500 NRS=0.2500 M10 2 6 1 1 p L=2.0u W=25.4u PD=35.4u PS=2.0u AD=127.2p AS=25.4p +NRD=0.1965 NRS=0.0393 M11 3 6 1 1 p L=2.0u W=12.3u PD=22.3u PS=2.0u AD=61.6p AS=12.3p +NRD=0.4061 NRS=0.0812 vdd 1 0 dc 5v vreset 5 0 pulse ( 5v 0v 2ns 0.2ns 0.2ns 1000ns 1000ns ) .tran 0.1ns 20ns .plot tran v(2) v(3) v(4) v(5) v(6) (-.2,5.2) .end ========================================================================= Steve Burns Caltech steveb@vlsi.cs.caltech.edu