Commit Graph

150 Commits

Author SHA1 Message Date
Jeff Young 6d296038f3 Improve simulation error reporting.
1) More REPORTER, less exception processing
2) Remove UI calls from SPICE_MODEL
3) Don't replace netlist with errors; show both
4) Don't bail out of netlist generation after single error

Fixes https://gitlab.com/kicad/code/kicad/issues/14295
2023-03-16 16:12:29 +00:00
jean-pierre charras 6343147bb3 Fix Coverity warnings. 2023-03-08 11:37:56 +01:00
Wayne Stambaugh 4e99812145 Coverity fixes and code cleaning. 2023-03-04 13:02:05 -05:00
Jeff Young 190ee58ccf Don't deliver pedantic error messages to user. 2023-03-02 00:38:39 +00:00
Jeff Young f5edcf82f2 Move only-write-model-if-it-contains-overrides up to base class.
It's necessary to keep us from writing out model lines for included
models and fall-backs of included models.
2023-03-01 13:31:57 +00:00
Jeff Young 1fdc81e68d Make substrate and thermal junction nodes optional.
Fixes https://gitlab.com/kicad/code/kicad/issues/14083
2023-02-25 20:36:30 +00:00
Jeff Young 2e5bf0210a Clear model name when switching to internal model.
Also fixes a bug to ensure that a modelLine gets written when using
an internal model.

Fixes https://gitlab.com/kicad/code/kicad/issues/14083
2023-02-25 15:22:20 +00:00
Jeff Young 9a0aeb4344 Adjust AC-small-signal plot names when saving placeholders.
Fixes https://gitlab.com/kicad/code/kicad/issues/14016
2023-02-25 13:26:33 +00:00
Jeff Young 5a12e5ee76 Only write included models if they have overridden parameters.
Fixes https://gitlab.com/kicad/code/kicad/issues/13953
2023-02-24 20:31:35 +00:00
Jeff Young 8059edff0e Don't mess up user formatting if eval'ing didn't actually change value.
Fixes https://gitlab.com/kicad/code/kicad/issues/13989
2023-02-24 12:44:55 +00:00
Jeff Young 71df3f9edb Don't allow extra text in Sim.* fields.
(In particular, the field name.)

Fixes https://gitlab.com/kicad/code/kicad/issues/13996
2023-02-24 12:44:55 +00:00
Jeff Young fe2679d965 Implement default levels for JFETs, MOSFETs and MESFETs. 2023-02-22 15:54:18 +00:00
Jeff Young cb71ea872b Don't allow 0 field text size.
Fixes https://gitlab.com/kicad/code/kicad/issues/13987
2023-02-22 11:40:49 +00:00
Jeff Young 68fe146861 Remove most of SIM_VALUE in favour of good old wxString.
This allows us to save the user's intent, including units, formatting,
and crucially variable references.
2023-02-22 10:45:27 +00:00
Jeff Young 90adc26b73 Remove dead and redundant code. 2023-02-22 10:45:27 +00:00
Marek Roszko aacc9746e3 Yeet the last of wxSize/wxPoint to stop leaking gdicmn everywhere 2023-02-18 22:57:18 -05:00
Jeff Young d1f6ba77e8 Better error reporting, and don't infer models with library references. 2023-02-04 20:55:52 +00:00
Graham Keeth a51dc8c9de HiSIM_HV MOSFET models are *not* the default type
The ngspice manual does not indicate that there is any default
model type. This change ensures we get a `level=xy` written out to the
.model line when we select a HiSIM_HV model.
2023-02-04 00:03:25 +00:00
Graham Keeth 41c4ee2245 ngspice requires J prefix for JFETs in netlist
Fixes: https://gitlab.com/kicad/code/kicad/-/issues/13774
2023-02-04 00:03:25 +00:00
Jeff Young 8b03c093f9 Move potentiometer pin model to r0, wiper, r1, and remove flipping code.
Fixes https://gitlab.com/kicad/code/kicad/issues/13741
2023-02-02 16:22:13 +00:00
Jeff Young c939b1ef76 Followed-by-3-digits doesn't guarantee a thousands separator.
Fixes https://gitlab.com/kicad/code/kicad/issues/13708
2023-01-30 21:22:48 +00:00
Seth Hillbrand 00f93db2ff Convert wide chars before reading model
The PEGTL parser used by KiCad for SPICE models only handles ASCII
characters.  Despite the docs stating that you can, in theory, parse a
UTF-8 character by reading the base string, in pratice this does not
work as the UTF-8 string is represented by unsigned chars and the PEGTL
string is based on signed char.

To work around this, we need to convert micro and mu to the ASCII 'u'
equivalent

Fixes https://gitlab.com/kicad/code/kicad/issues/13642
2023-01-23 15:23:41 -08:00
Jeff Young 959e20f52c A fall-back model for FOO should supply FOO's pin names.
Fixes https://gitlab.com/kicad/code/kicad/issues/13589
2023-01-19 12:14:50 +00:00
Jeff Young d839f11d49 Rip out parsing of SPICE libraries for netlisting.
We only need to parse the libraries for the Sim Model Editor dialog
(so that we can determine the models to put in the popup).  Doing it
for netlisting just opens us up to incorrectly parsing the SPICE, and
returns no value.

This also means that a SIM_LIB_MGR manages a single library, and
never multiple libraries.

This also allows the tidying-up of some error reporting structures to
better inform the user what went wrong.

Fixes https://gitlab.com/kicad/code/kicad/issues/13431
2023-01-12 14:51:03 +00:00
Jeff Young c43050b91b Remove safety around reading default parameter from Value field.
This should be OK now because we should be writing out all parameters
now (even if they're default value).

But perhaps more to the point, if we have the safety in there then
we miss cases where we really can't parse the model (value of "{VCC}")
and we need to fall back to a raw spice model.

Fixes https://gitlab.com/kicad/code/kicad/issues/13444
2023-01-09 15:24:45 +00:00
Jeff Young d65cf1b457 Check parameters before assuming we can handle the model.
Fixes https://gitlab.com/kicad/code/kicad/issues/13444
2023-01-08 18:00:22 +00:00
Jeff Young 7af718f1a5 Don't use std::string for things with wxString APIs.
It's too bug-prone.

Fixes https://gitlab.com/kicad/code/kicad/issues/13380
2023-01-05 14:16:39 +00:00
Seth Hillbrand cbe6cd3048 Don't move local vars
Copy ellision allows optimization if we don't explicitly move a local
variable in the return
2023-01-03 10:20:18 -08:00
Jeff Young baa13b885f Don't add Sim.Params unless it contains data, and update sallen key demo. 2023-01-03 17:20:23 +00:00
Jeff Young 9583b28063 Fix store-in-value bugs.
1) Always honour the dialog checkbox value.
2) Don't expect the absence of the primary parameter in Sim.Params to
   mean that Value stores it.  It might be the default parameter value.
3) Update spiceTypeInfo when matching legacy data to an internal model.
4) Initialize the store-in-value checkbox depending on whether or not
   we found the primary parameter in the Value field.
2023-01-03 17:20:23 +00:00
Jeff Young 9b9795a87d Reduce reliance on exception processing -- it's waay too bugy at present.
This moves some stuff to REPORTER APIs.  Moving more stuff would be good,
but it probably too high-risk at present.  We'll wait for 8.0 for that....

Fixes https://gitlab.com/kicad/code/kicad/issues/13359
2023-01-03 17:20:23 +00:00
Jeff Young fdc8f9c59d Fix accidental deletion in "TEXT_ATTRIBUTES" commit. 2023-01-03 01:23:18 +00:00
Jeff Young 7f35323416 Re-enable sim model ERC testing. 2023-01-03 00:59:14 +00:00
Jeff Young 277cd73204 TEXT_ATTRIBUTES doesn't hold the position.
Fixes https://gitlab.com/kicad/code/kicad/issues/13350
2023-01-02 17:36:21 +00:00
jean-pierre charras e8114dc49a Minor compil and Coverity warnings fixes. 2023-01-02 10:21:42 +01:00
Jeff Young 55b8f0bc00 Gracefully fall back to raw spice when we can't parse the data.
Fixes https://gitlab.com/kicad/code/kicad/issues/13329
2022-12-31 18:52:26 +00:00
Jeff Young 77b361a077 Add reference material to SPICE model pin assignment tab.
Fixes https://gitlab.com/kicad/code/kicad/issues/13309
2022-12-31 18:52:26 +00:00
Jeff Young 5deee405da Fill in built-in-model-type (which might not be the same as initial spice type).
Fixes https://gitlab.com/kicad/code/kicad/issues/13324
2022-12-31 01:07:46 +00:00
Jeff Young 62bec69d8c Migrate field text attrs from legacy fields.
Fixes https://gitlab.com/kicad/code/kicad/issues/13323
2022-12-30 23:41:46 +00:00
Jeff Young 424cdf6699 Overflow safety.
Sentry issue https://sentry.io/organizations/kicad/issues/3419671947/events/11212297e66c47c7af7144b5f075e88c/
2022-12-26 20:33:04 +00:00
Jeff Young 3f49fa5361 Don't write sim.params for library model unless we actually have overrides. 2022-12-23 12:46:18 +00:00
Jeff Young 7f60b87d1d Implement fallback models for SPICE .lib models that we can't parse. 2022-12-22 22:35:09 +00:00
Jeff Young 7abfa46531 Inferred sim value improvements, and a unit test for them. 2022-12-21 17:03:49 +00:00
Jeff Young e64852c833 Clearer naming. 2022-12-21 17:03:46 +00:00
Jeff Young e7e3f3bced SIM_MODEL code simplification.
1) Reduce API surface.
2) Prefer for(int) loops when what we're looking for is the index of
   the item to be found
3) Move "_" escape processing to SIM_MODEL_NGSPICE (which is the only
   model which uses the escape mechanism)
2022-12-21 17:01:58 +00:00
Jeff Young 8db015bc0f Accept thousands separators in inferred values. 2022-12-19 20:05:54 +00:00
Jeff Young 58ce0bb804 Simulation bug fixes.
1) Use C-locale when converting output of parsers to strings.

2) InferSimModel() must run on local copy of fields, not the fields in
the symbol (which might be different if they've already been edited in
the Symbol Properties Dialog, for instance).

3) InferSimModel() should accept a deviceType already set to RLC.

4) Don't output trailing decimal separators if there's no fraction.
2022-12-19 18:14:13 +00:00
Jeff Young a3a2e2e5b1 Update demos and legacy qa files SpiceMapping -> Spice_Node_Sequence.
Also removes migration in the code as it will no longer be needed.
2022-12-19 12:04:53 +00:00
Jeff Young 04f50ce687 Generate default pinMap from symbol's pins when we don't know SIM_MODEL's pins. 2022-12-19 11:40:52 +00:00
Jeff Young df8eb59854 Don't use SPICE_MODEL_SUBCKT::CreatePins to crate a default pinMap.
(It creates a fully-unconnected map on the premise that it's easier
to add a pinMap than to edit one.  This is probably no longer true
now that you don't have to un-assign a pin before re-assigning it.)
2022-12-19 11:40:52 +00:00