01
Current-carrying layout
Copper-weight and pour geometry sized to IPC-2152 with real margin — validated, not chart-guessed.
Boards that won’t melt.
Up to 200 A continuous. Heavy copper, IMS and aluminum-core, bus-bar attach, GaN/SiC. Every high-power board ships with a TRM thermal report — because “looks fine” isn’t a spec.
200 Acontinuous handled
3 oz +copper / IMS
±3 °Csim vs. measured
GaN/SiCtopologies
The deliverables.
02
TRM thermal report
Steady-state + transient thermal of the whole board at worst-case ambient. Named deliverable, every time.
03
PDN + IR-drop
Voltage-drop analysis across the power path so the rail still regulates at full load.
04
Mechanical + thermal interface
Heat-sink, cold-plate, and bus-bar attach detailed for assembly, with TIM and torque callouts.
What we’ve built.
Topologies
- Buck / boost / buck-boost
- LLC + phase-shift full bridge
- GaN + SiC half/full bridge
- Multiphase VRM
- PoE / PoE++ injectors
Construction
- 2–4 oz heavy copper
- IMS / aluminum-core MCPCB
- Bus-bar + busbar-on-board
- Embedded copper coins
- Thick-film + AlN substrates
Applications
- EV onboard chargers + DC-DC
- Industrial motor drives
- Telecom rectifiers
- Battery management (BMS)
- Solar / energy storage
120 A EV buck stage, re-spun before the melt.
The brief
An EV onboard charger’s 120 A buck stage was running hot in NPI. The customer wanted to know if 2 oz copper plus heatsinks would survive a Phoenix summer.
What we found
- TRM predicted 96 °C on the inductor pad at 2 oz / 55 °C ambient — over the cap.
- The VRM-to-inductor copper necked down at a routing channel, creating a current bottleneck.
- No thermal relief plan for the high-side FET to the chassis.
Outcome
Re-spec’d to 3 oz copper, widened the bottleneck 60%, added a copper coin under the FET. TRM dropped the hotspot to 82 °C; bench measured 79 °C. Saved an NPI re-spin.
Common questions.
Do you guarantee the thermal numbers?
We deliver a TRM simulation correlated to within ~3 °C of bench measurement on past boards. It’s an engineering estimate with stated assumptions, not a warranty — but it’s the same tool that’s caught real melt-downs pre-fab. See PhySignoff.
Heavy copper or IMS — which?
Depends on power density and mechanical needs. We’ll model both in TRM and show you the tradeoff in the quote.
Can you do the magnetics layout too?
Yes — planar magnetics, custom inductor footprints, and winding-on-PCB are in scope.
Prove it won’t melt.
Every high-power board ships with a TRM thermal report. Fixed-fee band in 60 seconds.