Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
The Spice algorithm (and Mike Ehglhart's LTSpice guru's enhancements) try to speed up the calculations by stepping over times steps where it thinks not much is going on. Some times the time step guesstimate fails, and the algorithm needs some guidance.
Alec's example poses a problem for the speed-up guess. He provides a 1mV change on top of a 3V level. The algorithm sees a node V(b) where relative to 3V, not much is changing, and thinks it can speed up by changing the time step to one that is large enough to skip over the detail. Specifying a max time step of 100ns force an evaluation of the circuit at least that often, so the detail of what is actually happening on node V(b) is revealed.
Thanks guys. I've still got a long way to go up the LTS learning hill. For some reason I nearly always forget to consider/limit the timestep. In this case I didn't think a 1k sinewave would bother LTS one jot.
I was hoping to create a sim .SUBCKT model including a similar setup of voltage sources, but clearly this would have the timestep issue. I know .tran directives within a model are ignored, so is there any other way of enforcing the timestep limit by something in the model itself?
Morning Eric. Thanks for the link. A useful reference.
When using LTS it feels like I'm a novice pilot at the controls of the Starship Enterprise. All those switches, knobs and levers but which one does what?
I feel the same way, its an interesting learning curve and I am still on it.
I find checking over other peoples LTS post solutions very informative, they often have a more novel way of solving a problem.