Decision

ESH10000690 R0 (Mini-PSU Precision Module) is built on the exact same PCB as ESH10000662 (Mini-PSU). The two are component-value variants of one layout:

• ESH10000662 — tuned for power (24 V / ~2.5 A use case).
• ESH10000690 — tuned for current precision (light-table control).

The functional difference is set mostly by resistor values (current-sense shunts / control scaling), not by layout changes.

Rationale

The customer wants a slimmer, precision-focused variant for the Collimator LED Control Box (ESH10000672 R1). The light-table role therefore uses ESH10000690 R0, and ESH10000662 is intentionally NOT a dependency of ESH10000672 R1.

Rejected alternative

Use ESH10000662 directly in the 672 stack — rejected: its power-oriented current range/resolution is unsuitable for light-table intensity control (cf. R0 ISS-006, where 662's range/resolution was the core problem).

Follow-up

• When the ESH10000690 project is fleshed out, baseline it from the ESH10000662 design and capture the precise resistor delta (shunt values + control scaling) as the variant definition.
• See related_to link ESH10000690 → ESH10000662.

Decision

ESH10000690 R0 precision-variant electrical spec (vs ESH10000662 power variant):

• Max continuous output current: 600 mA per channel (reduced from 662's 1.5 A/ch).
• Current set (CC limit) resolution: 12-bit DAC control.
• Current readback (ADC measure) accuracy: ≤ ±0.1 mA, stable across the 0–600 mA range.
• Output voltage range unchanged: 0–24 V/ch (shared PCB, voltage path retained).

Rationale

The light-table current-source role needs fine, precise current control at low currents; 662's 1.5 A range/resolution was unsuitable (cf. R0 ISS-006). Lowering FS to 600 mA and adding a precise readback path is the variant's whole purpose. Difference realised mostly via current-sense shunt / control-scaling resistor values on the shared layout (see E-001).

Impact on requirements

Drove REQ-002 (600 mA), and updated load conditions in REQ-003/004/005/008. Added REQ-009 (readback accuracy ≤0.1 mA). REQ-006 (CC linearity + 12-bit resolution) pending LSB resolution — see related notice.

Follow-up

• Capture the exact resistor delta (shunt values + scaling) as the variant definition once the design is baselined from 662.

Decision

Use RIS = 150 mΩ for the LT3942 current-sense shunt on ESH10000690 (precision variant), vs 50 mΩ on ESH10000662.

Rationale

Maps the LT3942 linear CTRL window (0.25–1.15 V) to the 600 mA full-scale target: I_IS(MAX) = (VCTRL − 0.25 V) / (10 × RIS); at VCTRL = 1.15 V → RIS = 0.90 V / (10 × 0.6 A) = 0.15 Ω.

• 662 R1: RIS = 50 mΩ → 0.25–1.15 V maps to 0 → 1.80 A.
• 690 R0: RIS = 150 mΩ → 0.25–1.15 V maps to 0 → 600 mA.

This is the primary current-sense resistor delta of the variant (see E-001/E-002).

To confirm

⚠️ 150 mΩ is derived from the datasheet formula, not yet verified against the ESH10000690 schematic/BOM. Confirm the actual fitted shunt value when the design is baselined from 662. Component-value verification (gain/offset 0.5/0.25 → recompute for 150 mΩ) also pending.

Observation

Proposed current-set resolution of 0.125 mA/step is not achievable with a true 12-bit DAC at 600 mA full scale:

• 12-bit over 0–600 mA FS → LSB = 600/4096 = 0.1465 mA/step.
• If only a fraction of the 0–5 V DAC window is usable (as on 662, ~3571 active steps) → LSB ≈ 0.168 mA/step (coarser).
• 0.125 mA/step at 12-bit implies FS ≈ 512 mA (below the 600 mA max), or requires >12-bit resolution.

Why it matters

Blocks finalising REQ-006 (CC-limit linearity + set resolution). Need engineer decision on which target to hold:

1. Accept ~0.146 mA LSB at 600 mA FS / full 12-bit, or
2. Lower FS (e.g. 512 mA) to hit 0.125 mA LSB, or
3. Use higher-resolution set DAC.

Next step

Resolve target, then create REQ-006 with the agreed resolution + linearity (FS%) spec.