Francesco Lannutti

But from MNA point of view, if you generate the correct current branch, it should work... At the end, the expanded inner nodes from the models make the full circuit flat. So the matrix solver will see a single whole matrix, which includes those nodes. Are we sure this is not a pure convergence issue due to a peculiarity of those circuits? And then maybe the KCL Verification could solve? Thank you, Fra

Hi Holger, I think it's that one, but it's a 10years old work, so I don't remember exactly! :) It's part of my PhD, attached, along with LnL Separation (I think there is a branch for this too). Thank you, Fra

Hi Holger, no, no, because it was experimental for some models. It needs code model modification, like KLU... But we could brainstorm offline among all of us. Thank you, Fra

Hi guys, what do you think if we also include my work on KCL Verification? It's basically on the same topic of this one. Maybe we can combine it... Thank you, Fra

I think the last one is this one: https://sourceforge.net/p/ngspice/ngspice/ci/new_kirchhoff-5/tree/ But I don't code NGSPICE since a while... That illustrates the work, I'm NOT saying it's ready to be used :) . You can have a look and continue from where I left ;) . Thank you, Fra

Hi Keith, you can look at my preliminary work on KCL Verification. In that branch, the convergence criterion for f(V) is given by Kirchhoff Current Law Verification. The branch is quite old, so I'm not sure it even contains a B source component, but you could take inputs from there... Thank you, Fra

Ciao Holger, nice update! :) I had a quick look and it seems order1 and 2 are the same for the moment, right? Also, I have this PDF (attached) where there is a description of the local truncation error methods. I didn't read it, but I think you can find it useful, or maybe you have already implemented what is written in there! ;) Let me know! Thank you, Fra

Hi Brian, thx a lot! :) Fra