This circuit has 8 floating nodes. From inspection, these nodes are believed to be floating:
DISCHARGE
GND
DISCHARGE_C
CHARGE
CHARGE_C
AVG
HIGH
LOW
Warning: Singular matrix. Check node AVG
Starting Gmin stepping.
Warning: Singular matrix. Check node AVG
Added gshunt=6.49641e-14
Fatal error: Timestep too small(5.11499e-18) at t=3.99199e-05
I’m running into an issue with QSPICE where I get the same error when I increase the simulation time or expand the circuit. To troubleshoot, I’ve updated my model and uploaded the new file
Despite the adjustments, the problem remains. Is this related to a limitation in QSPICE, or am I missing a configuration step?
In @bordodynov reply, he disabled “Fast (less accurate) math” option.
This allow Qspice to run simulation with QSPICE80.exe instead of QSPICE64.exe. Use the search in this forum if you need more information about QSPICE80 vs QSPICE64, and the general procedure to resolve timestep is too small. Using 80 bits math is one of a common trick to resolve this issue as high precision math is required.