Adom Custom Electronic Load
Public Made by Adomby adom
Two-layer digitally controlled electronic-load prototype. Compact R11 is under active electrical, cooling and probe-access review; not a manufacturing release.
| Name | Last updated |
|---|---|
| README.md |
Q1 and input diode selection review
Research checked 2026-10-13. This is a candidate screening, not a BOM replacement or hardware qualification. The three cooling concepts and Adom Fields 3D solver work remain separate, unfinished validation tasks.
Installed design
design/bom.csv and design/netlist.json identify Q1 as IXTP15N50L2, between LOAD_POS and SHUNT_P, and D1 as SBR10U45SP5-13, anode DUT_FUSED and cathode LOAD_POS. D1 is a series input protection diode, not an AC bridge rectifier. Current operating envelope is 4.5–24 V, min(3 A, 5 W/V). At this envelope the maximum commanded current is approximately 1.11 A; 3 A belongs to a future higher-power envelope.
Q1 shortlist
| Part | Why consider it | Decision boundary |
|---|---|---|
| IXTP15N50L2, current | Linear-rated TO-220; existing thermal model uses Rjc 0.42 K/W | Keep as baseline. A 500 V rating is far above the 24 V operating range, but does not by itself make the choice wrong. |
| IXTP80N075L2 | 75 V linear-rated TO-220; published Rjc 0.35 K/W | Worth checking for higher current / lower voltage compliance. Manufacturer-hosted sheet retrieved is marked Advance Technical Information: obtain current production specification before selection. |
| PSMN4R8-100YSE | Production 100 V enhanced-SOA ASFET, LFPAK56E, approximately 30 mm² footprint | Compact alternative, but requires a different cooling/PCB path. Hot-swap pulse capability is not proof of continuous electronic-load capability. Verify DC SOA at actual case temperature. |
The 0.42 to 0.35 K/W change saves only 0.385 °C at the existing guarded 5.5 W Q1 test. It cannot compensate for a poor clamp, insulator, airflow path, or heatsink. Neither low Rds(on) nor a large headline power rating establishes suitability for continuous linear operation. Gate capacitance, op-amp drive and loop stability must be checked before replacing Q1.
Sources: current Q1 datasheet, 75 V candidate datasheet, Nexperia candidate.
D1 and alternatives
The present SBR10U45SP5 is already a low-forward-drop 45 V, 10 A part. Its manufacturer specifies 0.47 V maximum at 10 A, 25 °C; that is not its exact drop at our roughly 1 A operating point. The datasheet's quoted board thermal resistances use 2 oz copper, so they must not be substituted for analysis of our 1 oz board. Existing project D1 screening uses 0.5133 W; it is a model input, not measured loss.
A conventional STPS10L60 provides 60 V margin but a TO-220 package; it offers no demonstrated compactness advantage and is not selected. Direct diode candidates require comparing Vf and reverse leakage at the same current and junction temperature, not headline numbers at different test conditions.
The most promising architecture alternative is an LM74700-Q1 ideal-diode controller plus an appropriately rated external N-channel MOSFET. Its typical forward-drop regulation is 20 mV: approximately 22 mW in the pass element at 1.11 A while regulating, plus controller and other losses. This is illustrative, not a guaranteed replacement loss; include hot Rds(on), regulation limits, startup, reverse hookup, capacitance, and transients. The controller starts near 3.9 V, compatible in principle with the present 4.5 V minimum but unsuitable to assume for a future near-zero-voltage load.
TI identifies LM74700D-Q1 as a newer alternative, but its SOIC package is larger than the original SOT-23 option. Newer does not automatically help the molecule footprint. A single ideal-diode FET is not a bidirectional disconnect: its body diode still permits forward conduction when the controller is disabled.
Sources: D1 datasheet, ST comparison, TI LM74700-Q1, TI newer SOIC alternative.
Recommendation and outstanding checks
Retain current Q1 and D1 as the reference while modeling the ideal-diode input as an alternate. It could substantially reduce the PCB diode hotspot and its copper-area demand, but adds parts, cost, and another control loop. At fixed DUT voltage/current the removed diode loss largely transfers into Q1: total DUT power still becomes heat. The alternate must be included in the Q1/heatsink model rather than claimed as a reduction in overall cooling requirements.
No claim is made that the installed pair is the newest or globally best. Supplier pricing and availability for alternatives have not been verified; existing BOM prices are dated snapshots. Before adoption: obtain final datasheets/lifecycle evidence, compare hot DC SOA over the full voltage/current envelope, verify current-control stability and startup/reverse faults, calculate footprint including supporting parts and cooling, and update libraries, sourcing, thermal analysis and tests. No schematic, PCB, firmware limit, or acceptance gate was changed by this review.