Repository navigation
Conversation
…rised colour mapping
solve_maxwell runs one Laplace solve per signal conductor (that conductor at 1 V, every other fixed conductor at 0 V, floats via float_groups) plus the same set in vacuum, and stores the energy matrices partitioned by material code. rlgc_matrices turns those into C, G and L_ext at any frequency via MaterialRecord.er_complex. Also fixes two normalization bugs in diff_modes that the N = 2 cross-check exposed: the mode capacitance was multiplied by pixel_width^2 (the pixel factors already cancel, as in cgp._energy_integral) and the odd/even drives were normalized by the differential drive voltage instead of per line, so z_odd was really Z_diff and z_even really Z_common. Both solvers now agree to 1e-11 on the same geometry.
… document conventions solve_maxwell now reports two previously silent conventions through MaxwellResult.warnings: conductor-material pixels with no conductor_ids entry are driven at 0 V (unlabelled_conductor_pixels), and the AMG Laplace path has no floating-conductor support, so a map with floats solved through AMG loses the equipotential constraint (float_groups_ignored_by_amg). Both are documented in the docstring, and the er field is now built by one helper for the nominal and the f_ref_hz branch. Also corrects the diff_modes module docstring to the post-fix meaning of its outputs (per-line odd/even impedances, Z_diff = 2 Z_odd, Z_common = Z_even/2), scopes the n_lines normalization comment to the two symmetric drives, and restores a zero-drive guard.
…nductance with roughness
…very skipped term
CHANGELOG.md: record every capability tasks 1-7 added (Usermap.from_labels, DSParams/RoughnessSpec/er_complex, solvers/roughness.py, solve_maxwell, rlgc_matrices, solve_wheeler_resistance, solve_impedance_matrix, SolverBudgetExceeded, to_touchstone_nport) and the diff_modes fix, with the exact old scale factors (z_odd off by 2/px**2, z_even off by 0.5/px**2) so anyone with numbers saved from the old bitmap differential path knows to re-run them. docs/theory/faraday_solver.md: add a short N-conductor extraction workflow section walking through labels, solve_maxwell, rlgc_matrices, solve_wheeler_resistance, roughness, solve_impedance_matrix and to_touchstone_nport in order.
- add a Touchstone version-2 two-port round-trip test (P=2, [Two-Port Data Order] 12_21) and a comment on the header line noting the nominal z_ref is for legacy readers, per-port values live in [Reference]. - add an absolute-magnitude cross-check: diff_modes.solve_modes' z_odd against sqrt((L11-L12)/(C11-C12)) from solve_maxwell + rlgc_matrices, the check the old internal-identity-only tests could not have caught. - document that MaterialRecord.resistivity_ohm_cm is in practice micro-ohm-centimetres and gets multiplied by 1e-8 for ohm-metres; the field keeps its name to avoid breaking the on-disk MoreColors.txt format. - move _CANNONBALL_HEIGHT/_CANNONBALL_RATIO into materials/database.py and import them from there in solvers/roughness.py instead of duplicating them: RoughnessSpec's validator can't call solvers.roughness directly without a cycle, since that module already imports RoughnessSpec. - add the from_labels validation tests and the float-role RGB test that were missing, closing the last gaps in that code's coverage.
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds what fieldforge's interconnect extraction needs from lineforge, as a library:
Usermap.from_labelswith integerconductor_ids(-1 non-conductor, 0 reference, 1..N signals, -2 floating) and a vectorised colour mapping.DSParams(Djordjevic-Sarkar) andRoughnessSpec(Hammerstad, Groisse, Huray cannonball) onMaterialRecord, wither_complex(f)and JSON pack I/O.solvers/mtl.py:solve_maxwell(N+1 Laplace solves) andrlgc_matrices(Maxwell C, L_ext = mu0 eps0 inv(C0), G from the dielectric energy shares).solvers/wheeler.py:solve_wheeler_resistance, the Wheeler incremental-inductance rule with roughness, giving the full R matrix.solvers/faraday.solve_impedance_matrix: N right-hand-side PEEC with one LU and aSolverBudgetExceededguard.touchstone.to_touchstone_nport: Touchstone 1 via scikit-rf and native Touchstone 2 with[Reference]and[Mixed-Mode Order].Also fixes a long-standing bug in
solvers/diff_modes.py: a spuriouspixel_width**2factor and per-drive normalisation made Z_odd 2/px^2 too large and Z_even 0.5/px^2 too small. Any externally saved bitmap differential-mode numbers from before this change are invalid; see the CHANGELOG entry.All existing tests unchanged; the five
tests/test_cli.pyfailures and the 17 mypy errors pre-date this branch.