[Kiwamu, Ellie, Aidan, Cao]
We noticed a fast degradation in the output power of the HWS-Y SLED over time. Despite the wrong calibration, calculation of the fractional power output shows that there is almost a 30% decrease:
| Time | Power (mean - 10 minute trend) (mW- WRONG calibration) |
| 1138577882 | 83.69 |
| 1138617346 | 79.66 |
| 1139790957 | 79.62 |
| 1140362164 | 70.25 |
| 1140830276 | 64.70 |
Between the 04/02 and 18/02, the SLED was off so the power output did not change. As soon as the SLED started being used again, the power output quickly decays. This means that if running continuously, the SLED's power output decayed 30% over 15 days. A plot from dataviewer is attached, which shows this clear decay.
This fractional 30% decay excluded the effect of amplication factor due to wrong calibration. The only effect of wrong calibration may have is some unknown offset that we are not aware of.
Nevertheless, it is worthwhile to investigate what is the cause of this fast degradation. We are suspecting whether the current driving the SLED exceed the limit but we are not aware of due to some calibration error. The current recorded from the channel SLEDCURRENTMON is around 100 mA, which is below the upper limit of this SLED. However, we will need to test this to confirm the actual current value.
It surprising that these SLEDs are degrading so fast. Last year Nutsinee and I adjusated the SLED driver boards (alog 18069, D1200600) so they can only output a maximum of 100mA, thinking that this would improve the lifetime of the SLEDs. The data sheet for these SLEDs indicates a maximum current draw of 350mA (see http://www.qphotonics.com/Fiber-coupled-superluminescent-diode-5mW-790nm.html and http://www.qphotonics.com/Fiber-coupled-superluminescent-diode-5mW-840nm.html). Yet it appears that limiting the max SLED current has not preventing the SLED power from dropping rather rapidly.