Solar MPPT Power System

@@ -141,6 +141,22 @@ of thumb:   on the same copper net, so it runs electrically in parallel and the pair is dominated   by the shorter one. Result: **92.0 → 20.2 mΩ**, with nothing existing moved. +**What that bought.** The same model, run on this board before and after those three+changes — "before" is what the copper measured when the model first flagged it, which is+what motivated each fix:++| Measurement | Before | After | |+|---|---:|---:|---|+| Panel input path resistance | 11.578 mΩ | **5.454 mΩ** | −53% |+| Panel input current density (p99) | 122.55 A/mm² | **55.24 A/mm²** | −55% |+| Buck-side switch node copper loss | 333.13 mW | **160.35 mW** | −52% |+| Battery bus → 5 V POL resistance | 76.856 mΩ | **30.297 mΩ** | −61% |+| Worst copper loss anywhere | 333.13 mW | **188.54 mW** | −43% |+| Peak current density, any case | 122.55 A/mm² | **61.29 A/mm²** | −50% |+| Peak board temperature | 48.5 °C | **45.7 °C** | −2.8 K |++Peak current density is halved, and no case on the board now exceeds 61.3 A/mm².+ The model also refuses to solve a case whose terminal has no copper reaching it, and reports it as unsolved rather than guessing — which is how an unrouted rail surfaces here before it surfaces in DRC.

@@ -41,6 +41,23 @@ between those two pads. Both ends land on the same copper net, so it runs electrically in parallel with the long one and parallel resistance is dominated by the smaller term. Nothing existing was moved. +**What that bought.** The same model, run on this board before and after those+three changes -- "before" is what the copper measured when the model first+flagged it, which is what motivated each fix:++| Measurement | Before | After | Change |+|---|---:|---:|---|+| Panel input path resistance | 11.578 mOhm | 5.454 mOhm | -53% |+| Panel input current density (p99) | 122.55 A/mm^2 | 55.24 A/mm^2 | -55% |+| Buck-side switch node copper loss | 333.13 mW | 160.35 mW | -52% |+| Battery bus -> 5V POL resistance | 76.856 mOhm | 30.297 mOhm | -61% |+| Worst copper loss anywhere | 333.13 mW | 188.54 mW | -43% |+| Peak current density, any case | 122.55 A/mm^2 | 61.29 A/mm^2 | -50% |+| Peak board temperature | 48.5 C | 45.7 C | -2.8 K |++Peak current density is halved, and no case on the board now exceeds+61.3 A/mm^2.+ This model also refuses to solve a case whose terminal has no copper reaching it, and reports it as unsolved rather than aborting the run. That is how unrouted rails surface here before they surface in DRC.

@@ -798,6 +798,23 @@ between those two pads. Both ends land on the same copper net, so it runs electrically in parallel with the long one and parallel resistance is dominated by the smaller term. Nothing existing was moved. +**What that bought.** The same model, run on this board before and after those+three changes -- "before" is what the copper measured when the model first+flagged it, which is what motivated each fix:++| Measurement | Before | After | Change |+|---|---:|---:|---|+| Panel input path resistance | 11.578 mOhm | 5.454 mOhm | -53% |+| Panel input current density (p99) | 122.55 A/mm^2 | 55.24 A/mm^2 | -55% |+| Buck-side switch node copper loss | 333.13 mW | 160.35 mW | -52% |+| Battery bus -> 5V POL resistance | 76.856 mOhm | 30.297 mOhm | -61% |+| Worst copper loss anywhere | 333.13 mW | 188.54 mW | -43% |+| Peak current density, any case | 122.55 A/mm^2 | 61.29 A/mm^2 | -50% |+| Peak board temperature | 48.5 C | 45.7 C | -2.8 K |++Peak current density is halved, and no case on the board now exceeds+61.3 A/mm^2.+ This model also refuses to solve a case whose terminal has no copper reaching it, and reports it as unsolved rather than aborting the run. That is how unrouted rails surface here before they surface in DRC.