We have a LS 218 controller which we are using to control a spectrometer at ~80 K. Two channels are actively controlling elements within the spectrometer and work very well at the ~1 mK level for many days. The other two channels (UNCONTROLLED) show the correct temperature ~79 K, but with a ~5mK oscillation with 8-10 minute period. There are no obvious heat sources which could be producing this. Also we swapped the inputs to these channels to other stable ones and these then started showing the same oscillation pattern.
Channela 5 and 6 always show an oscillating temperature no matter what cable or sensor is connected
All the thermometers are silicon diodes, DT-600 series (model DT-760A1-CU). I believe all the LS218 channels were setup identically, same sensor curves.