design-electromagnetic-device
About
This Claude Skill helps developers design practical electromagnetic devices like motors, generators, and transformers by bridging theory to application. It assists in sizing components for target fields or forces, selecting topologies, and computing performance metrics like torque and efficiency. The tool also analyzes key losses from copper resistance, core hysteresis, and eddy currents.
Quick Install
Claude Code
Recommendednpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/design-electromagnetic-deviceCopy and paste this command in Claude Code to install this skill
Documentation
Design Electromagnetic Device
Spec perf → topology → compute from EM first principles → analyze losses + efficiency → validate vs thermal + saturation.
Use When
- Size electromagnet (solenoid/toroidal) for B-field, pull, hold force
- Select motor topology (DC brushed, BLDC, stepper, induction), compute torque + speed + eff
- Design generator → V, I, freq
- Design transformer → V ratio, power, freq
- Analyze + min losses: copper (I^2 R), core (hyst + eddy), stray flux
In
- Required: Device type (electromagnet, motor, generator, transformer)
- Required: Perf reqs (B, force, torque, V ratio, power, eff target)
- Required: Operating (V, I, freq, duty, ambient T)
- Optional: Core mat (silicon steel, ferrite, powdered iron, air) + B-H
- Optional: Size/weight
- Optional: Cost/mfg
Do
Step 1: Reqs + operating
Full targets before topology:
-
Primary metric:
- Electromagnet: B (T) at point, or pull force (N) on armature
- Motor: rated T (N.m) at RPM, or power (W) at RPM
- Generator: V, I, Hz at mech speed
- Transformer: V1, V2, VA, freq
-
Secondary: Eff (%), max T rise above ambient (K), duty (cont, intermittent, pulsed), envelope (max D, L, weight).
-
Supply: V, I, freq (DC/AC w/ Hz), waveform (sine, PWM, trapezoidal).
-
Environment: T range, cooling (nat convection, forced air, liquid), altitude, vibration/shock.
## Design Requirements
- **Device type**: [electromagnet / motor / generator / transformer]
- **Primary specification**: [value with units]
- **Efficiency target**: [%]
- **Supply**: [voltage, current, frequency]
- **Thermal limit**: [max temperature rise in K]
- **Size constraint**: [dimensions or weight]
- **Duty cycle**: [continuous / intermittent (on-time/off-time) / pulsed]
→ Complete quantified reqs, no ambiguity. Every req has val + units.
If err: Conflict (high T in tiny vol + high eff) → identify tradeoff explicit. EM scaling: force ~ volume, losses ~ surface area, thermal constrains power density.
Step 2: Topology
Config matches reqs:
-
Electromagnet:
- Solenoid (cylindrical): Simple wind, uniform B = mu_0 n I. Uniform-field apps. Air gap for pull.
- Toroid: No stray field. Min stray. Less uniform for partial.
- C-core / E-core: High force compact. Air gap concentrates. Relays + hold magnets.
- Helmholtz pair: 2 coils sep by 1 radius. Uniform center. Calibration + measurement.
-
Motor:
- DC brushed: Simple drive, good low-speed T. Brushes limit lifetime + speed. T = k_T * I.
- BLDC: Electronic commutation, higher speed + lifetime. Trapezoidal/sinusoidal. Modern dominant.
- Stepper: Precise open-loop (1.8 or 0.9 deg). Lower cont T than BLDC. Positioning w/o feedback.
- AC induction: Robust, no PM, simple. Speed = supply freq + slip. Industrial power.
-
Generator: Motors reversed. BLDC motor → BLDC gen (back-EMF = output). Induction above sync. PM gen for small (wind, hydro).
-
Transformer:
- Core type: Windings on single leg. Std power.
- Shell type: Core around windings. Better shielding. High-power.
- Toroidal: No gap, low stray, compact. Higher winding cost. Audio + sensitive electronics.
- Planar / PCB: PCB trace windings. Low profile. SMPS at high freq.
## Topology Selection
- **Topology chosen**: [specific configuration]
- **Justification**: [why it matches the requirements]
- **Key advantages**: [for this application]
- **Key limitations**: [and mitigation strategy]
- **Alternatives considered**: [and why rejected]
→ Justified selection tied to Step 1 reqs w/ acknowledged limitations.
If err: No std topology meets all → hybrid (Halbach array) or relax secondary. Doc tradeoff.
Step 3: Design params
Physical dims + elec params from EM principles:
-
Electromagnet:
- Turns: N = B * l_core / (mu_0 * mu_r * I), or mag circuit: N * I = Phi * R_total
- Wire gauge: J (3-6 A/mm^2 cont, 15 A/mm^2 intermittent). A_wire = I / J.
- Core X-sec: A_core = Phi / B_max (below sat: 1.5-1.8 T silicon steel, 0.3-0.5 T ferrite)
- Gap force: F = B^2 * A_gap / (2 * mu_0) (Maxwell stress)
- R winding: R = rho_Cu * N * l_mean_turn / A_wire
-
Motor:
- Torque const: k_T = (2 * B * l * r * N) / phases (simplified BLDC)
- Back-EMF: k_E = k_T (SI)
- I_rated = T_rated / k_T
- omega_no_load = V_supply / k_E
- R from wire gauge + mean turn
-
Transformer:
- Turns ratio: N_1 / N_2 = V_1 / V_2
- Core X-sec: A_core = V_1 / (4.44 * f * N_1 * B_max) (sinusoidal)
- N_1 = V_1 / (4.44 * f * B_max * A_core)
- Window area: A_window = (N_1 * A_wire1 + N_2 * A_wire2) / k_fill (k_fill 0.3-0.5)
- Core vol: V_core = A_core * l_mean_path
-
Mag circuit (cores + gaps):
- R_core = l_core / (mu_0 * mu_r * A_core)
- R_gap = l_gap / (mu_0 * A_gap) (much > R_core for small gaps)
- R_total = R_core + R_gap (series), 1/R_total = sum(1/R_i) (parallel)
- Phi = N * I / R_total
## Design Parameters
- **Turns**: N = [value] (primary), N_2 = [value] (if applicable)
- **Wire gauge**: AWG [number] (diameter [mm], area [mm^2])
- **Core dimensions**: A_core = [mm^2], l_core = [mm], l_gap = [mm]
- **Core material**: [type], B_max = [T], mu_r = [value]
- **Winding resistance**: R = [Ohms]
- **Operating current**: I = [A], current density J = [A/mm^2]
- **Key performance**: [B-field / torque / voltage ratio = calculated value]
→ Numerical vals for all dims + elec params from EM equations w/ units at each step.
If err: Turns don't fit → bigger core (more window), finer wire (higher J, more heat), or reduce target. Core above B_max → bigger X-sec or more turns.
Step 4: Losses + eff
Quantify all mechanisms + eff:
-
Copper (I^2 R):
- P_Cu = I^2 * R_winding (DC)
- High freq: skin effect. R_AC / R_DC increases when diam > 2 * delta.
- Proximity effect in multi-layer → more AC R.
- Mitigate: Litz wire for >~10 kHz.
-
Core (hyst + eddy):
- Hyst vol per cycle: W_h = area B-H loop
- P_h = k_h * f * B_max^n * V_core (Steinmetz, n 1.6-2.0, k_h from data)
- Eddy: P_e = k_e * f^2 * B_max^2 * t^2 * V_core (t = lamination thick)
- Combined (gen Steinmetz): P_core = k * f^alpha * B_max^beta * V_core
- Mitigate: laminated (0.25-0.5 mm for 50/60 Hz, thinner higher freq), ferrite for >100 kHz
-
Eddy in conductors/structure:
- Stray flux → currents in frame, shields, conductors
- Big in large transformers + machines
- Mitigate: non-mag struct mat, mag shields
-
Mechanical (motors, gens):
- Bearing friction: P_friction = T_friction * omega
- Windage (air): P_windage ~ omega^3
- Brush friction (DC brushed): wear-dependent
-
Eff calc:
- Electromagnet: not primary metric → focus P = I^2 R for field/force
- Motor: eta = P_mech / P_elec = (T * omega) / (V * I)
- Generator: eta = P_elec / P_mech
- Transformer: eta = P_out / P_in = P_out / (P_out + P_Cu + P_core)
- Typ: small motors 60-85%, large 90-97%, transformers 95-99%
## Loss Analysis
| Loss Mechanism | Formula | Value (W) | Fraction of Total |
|---------------|---------|-----------|-------------------|
| Copper (I^2R) | [expression] | [W] | [%] |
| Core hysteresis | [expression] | [W] | [%] |
| Core eddy current | [expression] | [W] | [%] |
| Mechanical (if applicable) | [expression] | [W] | [%] |
| **Total losses** | | [W] | 100% |
- **Efficiency**: eta = [%]
- **Temperature rise estimate**: Delta_T = P_total / (h * A_surface) = [K]
→ Loss breakdown, each quantified, eff, T rise for thermal feas.
If err: Eff below target → ID dominant. Copper → bigger wire or fewer turns. Core → lower-loss mat or reduce B_max. Mech → better bearings. T rise exceeds → more cooling or reduce density.
Step 5: Validate
Meets specs + physically realizable:
-
Perf verify:
- Recompute primary from final params
- Meets/exceeds Step 1
- Margin: (achieved - required) / required %
-
Saturation:
- B_max < sat flux density of mat
- Every section (legs, yoke, gap fringing)
- Gap region lowest flux density, smallest X-sec has highest
-
Thermal:
- T_surface = T_ambient + P_total / (h * A_surface)
- Nat convection: h ~ 5-10 W/(m^2.K)
- Forced air: h ~ 25-100 W/(m^2.K)
- Insulation class: A (105°C), B (130°C), F (155°C), H (180°C)
- Core Curie: silicon steel ~770°C (rarely limit), ferrite ~200-300°C (can be limit)
-
Dims:
- Fits envelope
- Winding fits window w/ fill factor
- Clearance + creepage for HV
-
Margin + sensitivity:
- +/-10% var each key param (I, turns, gap, mu_r)
- ID most sensitive → drives mfg tolerance
- Air-gap: gap length almost always most sensitive
## Design Validation
| Requirement | Target | Achieved | Margin |
|------------|--------|----------|--------|
| [Primary metric] | [value] | [value] | [%] |
| Efficiency | [%] | [%] | [%] |
| Temperature rise | < [K] | [K] | [K margin] |
| Envelope | [dimensions] | [dimensions] | [fits / exceeds] |
## Sensitivity Analysis
| Parameter | Nominal | +10% Effect on Primary Metric | Most Sensitive? |
|-----------|---------|-------------------------------|----------------|
| Current | [A] | [+/- %] | [Yes/No] |
| Turns | [N] | [+/- %] | [Yes/No] |
| Air gap | [mm] | [+/- %] | [Yes/No] |
| mu_r | [value] | [+/- %] | [Yes/No] |
→ All reqs met w/ docs margins, thermal OK, most sensitive param ID'd.
If err: Req not met → iterate topology (Step 2), params (Step 3), or loss mitigation (Step 4). Thermal infeasible → reduce duty, more size (cooling), higher-T insulation, active cooling. Doc each iter.
Check
- All reqs quantified w/ vals + units
- Topology justified + alts docs
- Mag circuit complete (reluctances, flux, NI)
- Wire gauge for J (3-6 A/mm^2 cont)
- Core below sat w/ margin
- All losses quantified (copper, hyst, eddy, mech)
- Eff meets target
- T rise w/in insulation class
- Fits envelope
- Sensitivity ID's tightest-tolerance param
- Complete for prototype build
Traps
- Ignoring mag circuit reluctance: Gap dominates (1 mm gap > 100 mm silicon steel core reluctance). No mag circuit model → devices far below expectations.
- Operating above sat: Above B-H knee, incremental mu drops. Doubling I ≠ doubling flux. Appears to "stop working". Always check B_max in narrowest X-sec.
- Undersize copper for thermal: J limits = thermal in disguise. 10 A/mm^2 free air → overheats in min. Cont-duty < 5-6 A/mm^2 w/o active cooling.
- Neglect fringing at gaps: Flux spreads → effective gap area bigger. Gaps ~ core dim → fringing 20-50%. Ignoring → underestimate flux + overestimate NI.
- DC R at high freq: 10 kHz skin depth in Cu ~ 0.66 mm. Magnet wire > 1.3 mm diam → AC R >> DC R. Litz or parallel thin strands.
- Confuse k_T vs k_E units: k_T (N.m/A) + k_E (V.s/rad) numerically equal SI. BUT k_E in V/kRPM (datasheets) → convert: k_T [N.m/A] = k_E [V/kRPM] * 60 / (2 * pi * 1000).
→
analyze-magnetic-field— B-field from current dist for detailed analysissolve-electromagnetic-induction— induction principles in motors/gens/transformersformulate-maxwell-equations— full EM for high-freq, waveguides, antennassimulate-cpu-architecture— digital ctrl sys driving motor controllers + power electronicsanalyze-tensegrity-system— tension-compression networks; shares prestress equilibrium
GitHub Repository
Related Skills
content-collections
MetaThis skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.
polymarket
MetaThis skill enables developers to build applications with the Polymarket prediction markets platform, including API integration for trading and market data. It also provides real-time data streaming via WebSocket to monitor live trades and market activity. Use it for implementing trading strategies or creating tools that process live market updates.
creating-opencode-plugins
MetaThis skill helps developers create OpenCode plugins that hook into 25+ event types like commands, files, and LSP operations. It provides the plugin structure, event API specifications, and implementation patterns for JavaScript/TypeScript modules. Use it when you need to intercept, monitor, or extend the OpenCode AI assistant's lifecycle with custom event-driven logic.
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
