Menu
▾
▴
#821 Transient extra points
Hello, my weekly test for pre-master-46 has failed because there are two less data points in the transient simulation results. This is relative to my reference data (e.g. last week's results).
https://github.com/eelab-dev/EEcircuit-engine/actions/runs/19588752936/job/56102861604#step:12:32
Just making sure that this is intentional, not a result of a bug.
The netlist that I am testing is:
.include modelcard.CMOS90
r vdd 2 100.0
l vdd 2 1
c vdd 2 0.01
m1 2 1 0 0 N90 W=100.0u L=0.09u
vdd vdd 0 1.8
vin 1 0 0 pulse (0 1.8 0 0.1 0.1 15 30)
.tran 0.1 50
.end
This is not a bug.
With commit
b3eb6b12f ("Try to equalise the last two time steps before a breakpoint, if the second step would be smaller than CKTdelta otherwise.", 2025-11-15)
we have improved the time step generation, which may yield a lower number of time steps.
The number of time points is not a guaranteed value.
Thanks Holger, I will update my reference data. For reference, the commit is below:
https://github.com/danchitnis/ngspice-sf-mirror/commit/b3eb6b12fe39520b62b447ec06c89886db6e1ffc