Symbols for spice simulation need different pin numbering than used for connecting to footprints

[poeschlr]

Sadly there is (at the time of writing) no way to define different pin numbers for spice simulation and footprint connection. As ngspice seems to be inflexible with regard to pin numbering we might need to add specialiced symbols for simulation purposes.

Original discussion with more details: KiCad/kicad-library#182

Edit: Ideally this is fixed by introducing a new abstraction layer between symbols and simulation models. But this will have to wait till version 6. (Or even later)

[stambaughw]

Rene, There is no need to add special symbols for spice simulation. To change the spice node sequence to match the symbol pin out with the spice model, create a custom field "Spice_Node_Sequence" with the correct node sequence using integers separated with spaces. Cheers, Wayne

…

On 1/19/2018 9:19 PM, Rene Pöschl wrote: Sadly there is (at the time of writing) no way to define different pin numbers for spice simulation and footprint connection. As ngspice seems to be inflexible with regard to pin numbering we might need to add specialiced symbols for simulation purposes. Original discussion with more details: KiCad/kicad-library#182 <KiCad/kicad-library#182> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#189>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFnmpmypXhZhshyL1VRZc3tLE-PmUs5Aks5tMU0WgaJpZM4RlROk>.

[poeschlr]

@astroelectronica have a look at the information provided by @stambaughw. There is a symbol field that allows defining a different pin order for the spice export then used in the symbol.

[astroelectronica]

@stambaughw With some model files that is not possible because it's encrypted. In other cases, model file have not all pins, i.e., a MOSFET in SOIC-8 package, with several drain or source pins. The model just shows one connection for that.

Passives and semiconductors loss with a subcircuit (.subckt) associated. Circuit reference can't be C, L, Q, D, .... and all them turn to U.

I'm not sure if to modify simulation models is a good idea, because these files are verified by manufacturer.

Thanks.

[stambaughw]

On 1/21/2018 7:21 AM, astroelectronic@ wrote: @stambaughw <https://github.com/stambaughw> With some model files that is not possible because it's encrypted. In other cases, model file have not all pins, i.e., a MOSFET in SOIC-8 package, with several drain or source pins. The model just shows one connection for that.

Please do not include encrypted spice models with kicad symbols. This violates every ethical argument for using open source. If users want to use them for your own projects, that is fine but we should not be supporting them in our symbol libraries.

Passives and semiconductors loss with a subcircuit (.subckt) associated. Circuit reference can't be C, L, Q, D, .... and all them turn to U.

This behavior is defined by the spice standard. Any .subckt statement requires a U reference.

I'm not sure if to modify simulation models is a good idea, because these files are verified by manufacturer.

We should not modify the symbol file, we should use the node sequence file to map the symbol to the spice model. There may be some odd models with control inputs that require custom symbols but most of the spice models I use do not require custom symbols.

…

Thanks. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#189 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFnmpofxo9U6gpkmKo0TLM51T_jsUfYXks5tMyuvgaJpZM4RlROk>.

[evanshultz]

@poeschlr
Can this be closed? If so, please close it.

[poeschlr]

I think so.