gmoverid

Name: gmoverid
Author: Arcadia-1

116stars22forks

How to install gmoverid?

arcadia-1/gmoverid-skill/gmoverid

$npx -y skills add Arcadia-1/gmoverid-skill --skill gmoverid

Installs into the current project.

Summary

gm/ID transistor characterization and design methodology, based on ngspice + Python. Two independent workflows: (1) Characterization — generates three standard curve sets for any MOSFET model: gate capacitance (Cgg/Cgs/Cgd/Cgb vs Vgs), gm/ID four-quadrant characteristics (gm/Id v

Files · 1

View on GitHub

SKILL.md

1	# gm/ID Characterization and Design Skill
2
3	> Important — do not modify skill files during normal use.
4	> All code edits, new scripts, plots, and simulation outputs should go into the user's project working directory (outside `.claude/`), not into this skill folder. Only modify the skill assets (`assets/`, `SKILL.md`, `references/`) when the user explicitly asks to improve or extend the skill itself.
5
6	Dependency: `ngspice` skill (netlist execution, wrdata parsing). Model files are built in — the `transistor-models` skill is not required.
7
8	## Asset Files
9
10	```
11	assets/
12	├── simulate_gmoverid.py — ngspice simulation engine, data extraction, analytical caps, MODEL_INFO registry
13	├── plot_gmoverid.py — all matplotlib plotting functions
14	├── run_gmoverid.py — 180 nm single-node orchestration (NMOS/PMOS + channel-length sweep)
15	├── run_multinode.py — multi-node orchestration (45/22 nm HP)
16	├── design_gmoverid.py — GmIdTable lookup/sizing API + print_op()
17	├── models/ — built-in PTM model files (ready to use)
18	│ ├── nmos180.lib — 180 nm BSIM3v3 (VDD = 1.8 V)
19	│ ├── pmos180.lib
20	│ ├── nmos45hp.lib — 45 nm HP BSIM4 (VDD = 1.0 V)
21	│ ├── pmos45hp.lib
22	│ ├── nmos22hp.lib — 22 nm HP BSIM4 (VDD = 0.8 V)
23	│ └── pmos22hp.lib
24	└── netlist/
25	├── gmoverid_vgs.cir.tmpl — NMOS Vgs sweep (fixed Vds)
26	├── gmoverid_vds.cir.tmpl — NMOS Vds sweep (fixed Vgs)
27	├── gmoverid_pmos_vsg.cir.tmpl — PMOS \|Vsg\| sweep (fixed \|Vsd\|)
28	└── gmoverid_pmos_vsd.cir.tmpl — PMOS \|Vsd\| sweep (fixed \|Vsg\|)
29	```
30
31	Built-in models are from [PTM — Arizona State University (ptm.asu.edu)](https://ptm.asu.edu), free for academic research. When citing, use: W. Zhao and Y. Cao, "New Generation of Predictive Technology Model for Sub-45 nm Early Design Exploration," IEEE Trans. Electron Devices, vol. 53, no. 11, pp. 2816–2823, Nov. 2006. doi: 10.1109/TED.2006.884077. For nodes beyond the three built-in ones, install the `transistor-models` skill (full PTM library, requires manual MODEL_INFO configuration) or download from [mec.umn.edu/ptm](https://mec.umn.edu/ptm) and copy the `.lib` file into the project's `models/` directory.
32
33	---
34
35	## Workflow 1: Characterization (Simulation + Plotting)
36
37	### Deployment
38
39	1. Copy all assets (including the `models/` subdirectory) to `<project>/`
40	2. Create empty directories `plots/` and `logs/`
41	3. Run `python run_gmoverid.py` (180 nm) or `python run_multinode.py` (HP multi-node)
42
43	For additional nodes, simply copy the extra `.lib` file into `<project>/models/`.
44
45	All paths resolve automatically via `Path(__file__).resolve().parent` — no path configuration needed.
46
47	### Three Standard Plot Sets (per model)
48
49	\| Set \| Function \| Output file \|
50	\|-----\|----------\|-------------\|
51	\| Gate capacitance \| `plot_caps()` \| `gmoverid_caps_{model}_L{node}nm.png` \|
52	\| gm/ID four-quadrant \| `plot_main()` \| `gmoverid_{model}_L{node}nm.png` \|
53	\| IV characteristics \| `plot_iv()` \| `gmoverid_iv_{model}_L{node}nm.png` \|
54
55	gm/ID four-quadrant layout (`plot_main`, 2×2):
56	```
57	(0,0) gm/Id [V⁻¹] vs Vov \| (0,1) Id/W [uA/um] vs gm/Id (log Y)
58	(1,0) fT [GHz] vs gm/Id \| (1,1) gm·ro vs gm/Id
59	```
60	- All four panels use Vds as the curve parameter (VDS_LIST)
61	- gm/Id X-axis: xlim = [4, 24], one grid division per 2 V⁻¹
62	- Panels (0,1), (1,0), (1,1) apply `_fall_mask(gmid)` before plotting — the gm/ID array is double-valued (same gm/ID appears on the subthreshold rising side and the inversion falling side); only the falling (inversion) side is kept to prevent curve fold-back. Panel (0,0) is exempt because its X-axis is Vov, which is monotonic.
63
64	### Adding a New Technology Node
65
66	1. Copy the model `.lib` into `models/` (download from [mec.umn.edu/ptm](https://mec.umn.edu/ptm) or install the `transistor-models` skill)
67	2. Add an entry to `MODEL_INFO` in `simulate_gmoverid.py` (see conventions.md §3)
68	3. Add an entry to `NODE_CFG` in `run_multinode.py`
69
70	---
71
72	## Plotting Conventi

People who liked this also viewed

6 picks

Type

gmoverid

116stars22forks

View on GitHub

How to install gmoverid?

arcadia-1/gmoverid-skill/gmoverid

$npx -y skills add Arcadia-1/gmoverid-skill --skill gmoverid

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# gm/ID Characterization and Design Skill
2
3	> Important — do not modify skill files during normal use.
4	> All code edits, new scripts, plots, and simulation outputs should go into the user's project working directory (outside `.claude/`), not into this skill folder. Only modify the skill assets (`assets/`, `SKILL.md`, `references/`) when the user explicitly asks to improve or extend the skill itself.
5
6	Dependency: `ngspice` skill (netlist execution, wrdata parsing). Model files are built in — the `transistor-models` skill is not required.
7
8	## Asset Files
9
10	```
11	assets/
12	├── simulate_gmoverid.py — ngspice simulation engine, data extraction, analytical caps, MODEL_INFO registry
13	├── plot_gmoverid.py — all matplotlib plotting functions
14	├── run_gmoverid.py — 180 nm single-node orchestration (NMOS/PMOS + channel-length sweep)
15	├── run_multinode.py — multi-node orchestration (45/22 nm HP)
16	├── design_gmoverid.py — GmIdTable lookup/sizing API + print_op()
17	├── models/ — built-in PTM model files (ready to use)
18	│ ├── nmos180.lib — 180 nm BSIM3v3 (VDD = 1.8 V)
19	│ ├── pmos180.lib
20	│ ├── nmos45hp.lib — 45 nm HP BSIM4 (VDD = 1.0 V)
21	│ ├── pmos45hp.lib
22	│ ├── nmos22hp.lib — 22 nm HP BSIM4 (VDD = 0.8 V)
23	│ └── pmos22hp.lib
24	└── netlist/
25	├── gmoverid_vgs.cir.tmpl — NMOS Vgs sweep (fixed Vds)
26	├── gmoverid_vds.cir.tmpl — NMOS Vds sweep (fixed Vgs)
27	├── gmoverid_pmos_vsg.cir.tmpl — PMOS \|Vsg\| sweep (fixed \|Vsd\|)
28	└── gmoverid_pmos_vsd.cir.tmpl — PMOS \|Vsd\| sweep (fixed \|Vsg\|)
29	```
30
31	Built-in models are from [PTM — Arizona State University (ptm.asu.edu)](https://ptm.asu.edu), free for academic research. When citing, use: W. Zhao and Y. Cao, "New Generation of Predictive Technology Model for Sub-45 nm Early Design Exploration," IEEE Trans. Electron Devices, vol. 53, no. 11, pp. 2816–2823, Nov. 2006. doi: 10.1109/TED.2006.884077. For nodes beyond the three built-in ones, install the `transistor-models` skill (full PTM library, requires manual MODEL_INFO configuration) or download from [mec.umn.edu/ptm](https://mec.umn.edu/ptm) and copy the `.lib` file into the project's `models/` directory.
32
33	---
34
35	## Workflow 1: Characterization (Simulation + Plotting)
36
37	### Deployment
38
39	1. Copy all assets (including the `models/` subdirectory) to `<project>/`
40	2. Create empty directories `plots/` and `logs/`
41	3. Run `python run_gmoverid.py` (180 nm) or `python run_multinode.py` (HP multi-node)
42
43	For additional nodes, simply copy the extra `.lib` file into `<project>/models/`.
44
45	All paths resolve automatically via `Path(__file__).resolve().parent` — no path configuration needed.
46
47	### Three Standard Plot Sets (per model)
48
49	\| Set \| Function \| Output file \|
50	\|-----\|----------\|-------------\|
51	\| Gate capacitance \| `plot_caps()` \| `gmoverid_caps_{model}_L{node}nm.png` \|
52	\| gm/ID four-quadrant \| `plot_main()` \| `gmoverid_{model}_L{node}nm.png` \|
53	\| IV characteristics \| `plot_iv()` \| `gmoverid_iv_{model}_L{node}nm.png` \|
54
55	gm/ID four-quadrant layout (`plot_main`, 2×2):
56	```
57	(0,0) gm/Id [V⁻¹] vs Vov \| (0,1) Id/W [uA/um] vs gm/Id (log Y)
58	(1,0) fT [GHz] vs gm/Id \| (1,1) gm·ro vs gm/Id
59	```
60	- All four panels use Vds as the curve parameter (VDS_LIST)
61	- gm/Id X-axis: xlim = [4, 24], one grid division per 2 V⁻¹
62	- Panels (0,1), (1,0), (1,1) apply `_fall_mask(gmid)` before plotting — the gm/ID array is double-valued (same gm/ID appears on the subthreshold rising side and the inversion falling side); only the falling (inversion) side is kept to prevent curve fold-back. Panel (0,0) is exempt because its X-axis is Vov, which is monotonic.
63
64	### Adding a New Technology Node
65
66	1. Copy the model `.lib` into `models/` (download from [mec.umn.edu/ptm](https://mec.umn.edu/ptm) or install the `transistor-models` skill)
67	2. Add an entry to `MODEL_INFO` in `simulate_gmoverid.py` (see conventions.md §3)
68	3. Add an entry to `NODE_CFG` in `run_multinode.py`
69
70	---
71
72	## Plotting Conventi

People who liked this also viewed

6 picks

Type