# Electron Microscope Theory #publish Some notes on how we think about [[Electron Microscope MOC|electron microscope]]s, optics and otherwise. - [[Electron Microscopes observe at scales far shorter than human perception, and exist on time scales much longer than human perception]] - Data loggers can prove much more useful than oscilloscopes. A lot of the time (captured) is a useful thing, most of the time. - Vacuum problems tend to happen over the course of 10-30 minutes. Trying to remember everything that happens in that time, every time you try to diagnose it, is asking a lot. - [[time series]] - [[Things we have, or have wanted to have, recorded with time series data collection]] - Why diffusion pumped microscopes sometimes go from "yeah theres a few finger prints" to "theres oil literally everywhere", in no particular order: - [[Water Chiller]] - you should expect this to at least experience an interruption at some point, for planning reasons. - [[Interlock]]s failing to detect that the water chiller has stopped cooling. This should also be expected to happen, unless you test, inspect, and replace appropriately. - Cooling interlocks can generally be throughly tested safely - Cooling interlocks generally shut things down pretty fast. Loosing the water jacket on a 400w+ heater that is the size of a bagel is... Going to result in things heating up past 50°C in a hurry. That is more than twice as hot as the water that should normally cool the interlock. - On many diffusion pumps, the thermal interlock is mounted on the water jacket(often just tubes) that are closest to the heater.