Keita K., Giacomo C., Richard M., Filiberto C., Kiwamu I.

The resonant RF PD [1][2] that has been used for the length sensing of the IMC, was found to have two broken amps on the circuit board. This PD is the one which showed a sudden big offset in the DC signal as reported in [3]. We further investigated and fixed it. After fixing we tested the DC path and no big offset was observed, which is good.


Symptom
 The differential output was insane. The positive pin showed -2.2 V and the negative one showed -0.8 V with respect to the ground when no light is on the diode, resulting in an offset of -1.4 V in total. Though this number is not really what we got yesterday [3], the symptom remains the same --- a big offset and somehow magically the PD responds to light.

The repair
 We removed the PD from the IOT2L table and brought it to the EE shop to have a close look. It seemed that
 (1) one of two amplifiers for the differential output (U3 : at the very last stage of the DC path) seemed blown, resulting in such unbalance at the outputs.
 (2) the trans-impedance amplifier (U2) was not closing the feedback loop at all. In the same trans-impedance loop there was a current buffer (U5) and this was also suspicious initially.

 Filiberto replaced U3 and U2 with a new AD8672 and OP27 respectively. After the repair, it started working OK. We have tested the DC path by illuminating the diode with a visible laser pointer. It responded to the light, which tells us not only the DC path is OK but also the diode is still alive. Note that U2 was an AD8675 before the repair and interestingly it looked it had been replaced once as the foot prints were clearly dirtier than everyone else.


[1] D1101124-v6 
[2] SN:1203396
[3] LHO alog #5219