Alastair, Corey, Nutsinee
First RTD/IR sens. alarm went red on the medm screen. A visual inspection at the controller box showed IR FAULT alarm which couldn't be untripped by simply turn the key back and forth. Before going out to swap the IR sensor box I discovered that the laser had actually stopped working almost two hours before the alarm went off. Meaning, nothing was heating up the viewport at the time IR sensor tripped. We followed Alastair's suggestion (with a permission from Mike) and inspected the viewport from inside the table anyway to make sure there's nothing obvious (eg. bug crawed in and got zapped). While taking down the IR sensor box the sensor happened to untripped itself. So we swapped the DB9 cable and plug it to the spare IR box and left it on the chassis. After the IR fault got untripped we were able to restart the laser.
The cause of CO2X tripped is unclear. There was a sharp drop in power supply but only by a tiny amount (-0.005 V) at the same time that the laser output power, laser temperature, and the current drawn dropped to 0. All these happened almost two hours before IR FAULT alarm went off. The chiller was working fine.
Is the IR sensor output available to be plotted? Along with an output of a nearby temperature sensor? The comparator circuit in the IR sensor may have tripped due to a local rise in temperature, or an instability in its power supply voltages.
Unfortunately I don't think there's a channel that allow us to monitor the IR sensor readout. If the comparator box were to trip due to a local rise in temperature it should have happened sooner since the box has never been moved since we installed it.