narayan@cssun.tamu.edu (Sriram Narayan) (12/22/89)
I would like to report a few problems I have encountered using Spice2G.6 on our HP9000/300 series machines with HP-UX 6.5. Firstly, I would like to thank those who responded to my initial post for running this on HP machines. 1. The results I got for a "badly designed" CMOS circuit did not agree with the results I got with Spice running on a VAX and the results obtained from Spice3B.1. The VAX and 3B.1 results agreed. By "badly designed" I mean transistors in or around cutoff when they should be in saturation. This is only to show that one should be careful about interpreting results when your circuit is in a marginally biased condition. Has anyone noticed this problem? 2. We were simulating a circuit with noise analysis and at one point the program crashed, after giving a segmentation fault (core dumped) error. This was after DC convergence was obtained and part of the noise analysis was completed. This same circuit ran on the VAX without any problems. Comparison could not be made with the Spice3B.1 simulation since it doesn't have a noise analysis or transfer function option. If anybody desires further details or if they feel my results are due to improper installation of Spice2G.6 I would like to heare from you. Thanks. -Sriram ----------------------------------------------------------------------------- Sriram Narayan 810446 Texas A&M University | INTERNET: narayan@photon.tamu.edu | Go Alain Prost! US Mail: X4 Hensel #K, College Station, TX 77840.|