⏳ RUNNING - Simulation with tuned recombination parameters
Started: 2026-01-07 19:09:58
Elapsed: ~13 minutes
Expected: 15-20 minutes total (slower due to increased recombination)
# database.py - Silicon
'TAUN0': 0.1E-6, # 100 ns
'TAUP0': 0.1E-6, # 100 nsResult: Simulated current ~143x too high
# database.py - Silicon
'TAUN0': 1.0E-9, # 1 ns (100x reduction)
'TAUP0': 1.0E-9, # 1 ns (100x reduction)Expected: Simulated current should match experimental within ~2-10x
Physics:
- Shorter carrier lifetime → more recombination
- Shorter diffusion length (L ∝ √τ)
- Lower saturation current density Js
- Current reduces by factor of √(τ_old/τ_new) ≈ √100 ≈ 10x
Expected Results:
| Voltage | Experimental | Before (τ=100ns) | After (τ=1ns) Expected |
|---|---|---|---|
| 0.2V | 2.2 nA | 1.8 nA (0.82x) | ~0.2 nA (0.1x) |
| 0.4V | 49.4 nA | 7.06 µA (143x) | ~700 nA (14x) ✓ |
| 0.6V | 1.29 µA | 71.9 µA (56x) | ~7 µA (5.6x) ✓ |
| 0.8V | 38.6 µA | 56.7 mA (1466x) | ~5.7 mA (147x) |
The voltage-dependent error suggests we still need series resistance modeling (Rs ≈ 5Ω from experimental data).
- ✅ Wait for simulation to complete
- Check validation metrics (R², MAPE)
- If needed: Add series resistance correction
- Generate final comparison plots
Last updated: 2026-01-07 19:23