200 A Regen Steady-State Thermal Test — Gen7 Size 2

After the 450 A run tripped the inverter overtemperature protection in 78 s with the dry-mounted interface, the power stage was remounted with thermal paste. This run validates that rework: a 54.7-minute continuous hold at a 200 A q-axis command — long enough for the baseplate to reach true steady state — while a dynamometer-spun PMSM provided the mechanical drive. Because the machine ran at only ~470 RPM, the regen produced about 1 kW at a ~145 V bus: full 200 A of phase-current stress at very low power, in the spirit of the 100 A endurance test but now with a proper thermal interface.

Result: baseplate TS1 plateaued at 44.4 °C and TS2 at 44.0 °C against a 22.5 °C ambient (ΔT ≈ 22 K), with no faults during the hold. For comparison, the 200 A run on the previous dry, gapped interface was still climbing through 79 °C after 9 minutes at the same phase current.

Test setup

• DUT: OpenVVVF Gen7 size 2 (C2) power stage, remounted to its heatsink/plate with thermal paste; firmware foc_demo graph (hash 855dbcee90d611cb), OpenVVVF/RTE commit 5b529e2 ("fix gen7 current feedback and voltage control with bounded capture diagnostics").
• Machine: PMSM/IPM (FRAM config: 10 poles, R = 0.0121 Ω) coupled to the Sierra CP Engineering dynamometer at approximately −470 RPM mechanical (~−2,340 RPM electrical) during the hold.
• DC supply: Sorensen bench supply holding ~145 V on the bus. DC-link current stayed ~7 A, far below the supply's ~50 A regen sink limit, so no external load bank was needed.
• Instrumentation: RTE Studio telemetry, Klein Tools TI250 thermal imager (photos below, taken once the baseplate had reached steady state), and a ~2-minute bench video of the run including the thermal-imager passes (Steady-State-Run-Video.mp4, 1080p, transcoded from the phone original).

Test conditions

Procedure

1. Powered up; the gate driver latched a startup fault (with a software phase-overcurrent on the first start attempt) — both cleared by ~t = 27 s and control started at t = 43.3 s. See Observations.
2. Confirmed current orientation with small bidirectional jogs (IqVar 10/−20/30/−40 A).
3. Commanded IqVar = 200 at t = 112.75 s and held it for 54.7 minutes while logging telemetry.
4. Once the baseplate NTCs had plateaued (final ~20 minutes within ±0.1 °C), took TI250 thermal images and commanded IqVar = 0; logging continued through the cooldown.

Electrical results

The hold was electrically uneventful, which is the point of a steady-state thermal run:

• Current: 199.9 A average Iq, 208 A maximum, tight regulation throughout.
• Bus: 144.7 V average, 143.4–145.6 V — the supply held the bus with no clamp hardware in the loop.
• DC link: 6.6 A average, 949 W average (1.36 kW peak) — the low shaft speed limits back-EMF, so most of the phase-current work dissipates in the windings and switchgear rather than returning as bus power.
• Protection: gate_fault = 0 for the entire hold, cg_vlimit_scale = 1.000 throughout, no desaturation or overcurrent events after startup.
• Machine-terminal power (calculated): from the logged dq quantities (peak-amplitude convention, P = (3/2)(vd·id + vq·iq), Q = (3/2)(vq·id − vd·iq)), the machine delivered 0.94 kW mean while drawing ≈0.71 kvar of magnetizing reactive power (quoted as magnitude) — 1.18 kVA, power factor ≈ 0.80. Machine-terminal and DC-link power agree within ~1% — calculated inverter efficiency ≈ 100% at this ~1 kW level, where telemetry offsets dominate the difference.

Thermal results

The baseplate curve (telemetry overview below) is a clean single-pole rise to equilibrium:

• Baseplate TS1 (temp_inv1_c): 27.7 °C at the start of the hold → 43.3 °C at 17 min → plateau 44.4 °C (last 10 minutes average 44.32 °C, final sample 44.4 °C). Total rise ≈ 16.7 K above its starting point.
• Baseplate TS2 (temp_inv2_c): 30.8 °C → plateau 44.0 °C.
• Against the 22.5 °C operator-recorded ambient, steady-state ΔT is ≈ 21.9 K (TS1) at ~1 kW of returned power — and, more meaningfully, at 200 A of phase current, since inverter dissipation here is conduction-dominated and scales with current, not DC power.
• Comparison to the dry interface: the September 22 run at the same 200 A phase current (induction machine, ~1,100 RPM, gapped heatsink, no interface material) climbed from 34.5 °C past 79 °C in about 10 minutes and continued into the thermal OTP trip documented in OV-TEST-HW-THERMAL-OTP-200A-GEN7-SIZE2. The pasted assembly, by contrast, equilibrated ~22 K over ambient and could have held indefinitely. Machine and speed differ between the two runs, so this is a qualitative comparison, but the interface quality is the dominant difference.
• TI250 spot readings during the steady-state photo pass sat in the 34–42 °C range on the exterior surfaces (auto-scaled crosshair values, individual points rather than a surface survey):

Telemetry overview

Open full session: View the run in the Telemetry Viewer

Observations

• Startup faults (pre-test): on power-up the gate driver latched a fault (MAX22530 interrupt/filter reset complaints) and the first control start attempt also raised a software PhaseOvercurrent. Both cleared by ~t = 27 s and did not recur during the 55-minute hold. They are recorded here as boot-time transients on a fresh power-up, unrelated to the steady-state segment.
• Telemetry framing: this session's export splits telemetry into interleaved frames (a 49-key FOC bundle, a 2-key DC-link frame, and a 2-key encoder frame) rather than the single merged frame of earlier logs. The decimated log and plots in this report merge them by timestamp.
• Machine speed drooped slightly late in the hold (−489 → −447 RPM) without electrical consequence; DC-link power and current stayed nominal.
• TS3 (temp_inv3_c) remained disabled (wiring defect, unrelated to the test); temp_motor_c is unpopulated.

Conclusion

Pass. With the thermal-pasted interface, the Gen7 size 2 assembly holds a continuous 200 A regen command (~1 kW at ~470 RPM, ~145 V bus) at true steady state with baseplate temperatures of 44.4 / 44.0 °C — only ~22 K over the 22.5 °C ambient — versus an unbounded climb to the 80 °C protection threshold with the previous dry, gapped interface. The thermal rework is validated for continuous-duty operation at this current.

Artifacts

• Decimated telemetry log (JSONL, 1 sample/s, frames merged) — open in Telemetry Viewer
• Telemetry overview plot
• Session video (MP4, ~2 min, 1080p) — bench view of the run including the thermal-imager passes
• Full source telemetry log, retained for traceability
• Photos: bench setup, power stage detail, power stage on heatsink, control board detail, inverter side view, clamp hardware detail, chassis view
• Thermal images: view 1, view 2, view 3, view 4, view 5, view 6
• Full-resolution originals of the photos are retained alongside as *-Original.jpg.