With Mark,

During last week when Mark was testing new Guardian code, he found that SRM GAS BIO WD tripped, suggesting that the actuators might be satuated by the DC/DAMP control. He reset the suspension, cleared the controller buffers and managed to put the suspension into the GAS_CONTROL_ENGAGED guardian state.

Today, I tried to recreate the event by putting the SRM into SAFE state and then GAS_CONTROL_ENGAGED state. When the suspension is in GAS_CONTROL_ENGAGED state, the BF DC control output kept increasing at a fast rate even when the BF is around the setpoint. This might the the cause of the BIO WD trip.

In this case, the DC control was pulling the BF down from 6 µm to 0 µm. I attempted to put the BF further down, and I couldn't. This was also reported in 8672, 8733 and 8799 But, it can be actuated upward with seemingly no problem. This suggests that either the keystone or the IM is blocked by something below. To distinguish, I put ±500 µm offset on the F1 using the setpoint. If the IM is touching something below, there should be some observable increase in IM/BF vertical when the F1 is put at -500 µm but there isn't. Meanwhile, when the F1 is pulled to +500 µm, the BF keystone still couldn't be put further down, indicating that the one or more of the lock screws might be very close to the keystone. This could possibly be the case because we could have previously lock the keystone at 0 µm but forgot to release it or didn't release it well enough. 

One solution which will not involve opening the tank is to redefine the BF height to a higher position which would allow some margin for the BF and IM. But, we will need to discuss with other people to decide what should happen.