Unfortunately after swapping out the 16bit-DAC card in h1oaf0 almost two weeks ago, yesterday we got another zeroed-dac event. At 11:53 PDT, Sunday 25th September, the h1iopoaf0 model recorded an ADC, DAC and DK error in its STATE_WORD and from then onwards was correctly outputting zeros to all DAC channels. As usual, this was instructing the TCS chillers to drive to a low water temperature, and 52 minutes later the chiller tripped the TCS lasers (12:45 PDT).
Because we received both an ADC and DAC error, we could try swapping the first and second ADC cards.
T.J. has made his EPICS alarm announcement for this event more insistent, repeating every 2 minutes. If the control room can restart the models on h1oaf0 in time, we think the laser trips could be averted.