We aligned the ETMX OSEMs and ran into this issue. Looking at the SENSOR_SIDE channel, we pulled out the OSEM and determined that the open light voltage is 874 counts, so we centered it around 440 as well as we could. This is same channel as its slow counterpart SDSEN_OUTPUT (grey number immediately to the right on SUS medms).
The slow signal from the side sensor on ETMX was last seen in action sometime in May 2010! And then the frame builder has no data for a while on this channel. After that the channel shows some bistability starting Sept 2010 but has not been working. The fast channel of this sensor (C1:SUS-ETMX_SDSEN_OUTPUT) does work so the sensor is working. Probably is a loose contact... needs to be fixed.