Specially in cases where the furnace controller is guided by a thermocouple that is located inside the ProboStat, it is possible to get overshoot or oscillation of temperature.

The cause for both is delay. Heat takes time to penetrate from the furnace heating elements all the way to the thermocouple tip inside the sample holder (there are dense alumina walls to penetrate). This effect is particularly strong at low temperatures (below 400ºC). By default, the controller does not understand this delay and will try to heat more and more. This will cause the system eventually to heat more than what was required, and then correct for that overheating by under-heating and so on. It means that the temperature may oscillate around the target setpoint for quite a while. Each cycle can last even an hour as there is no active cooling.

To reduce this problem one must use a feature called ramp rate (on the controller), with slow ramp (for example 5 degree per minute). Ramp rate is a gradually moving setpoint between the current temperature and the target temperature. Using ramp rate will not eliminate the problem but it can make it much smaller.

Another possibility is to perform something called optimization of the PID settings. This is only recommended for users who have good understanding of the controllers inner workings. Good side of this solution is that it will almost completely eliminate the overshoot and oscillation, but the downside is that the process is difficult to perform, and will only solve the problem for a specific situation (Specific sample holder vertical position and a limited temperature range). For example if PID settings are optimized for "Internal thermocouple at sample holder middle of furnace at 400ºC" The same settings will not work very well at 800ºC, or not at all if the furnaces own thermocouple is used.

Perhaps best solution is to use the thermocouple of the furnace itself to control the furnace temperature. This solution will mostly eliminate overshoot and oscillation. The downside of this approach is that the furnace temperature and the sample temperature may not be the same. The difference depends on the general temperature and the location of the sample holder in the furnace. If exact sample temperature is needed, then read the sample holder thermocouple with another instrument, and adjust the furnace temperature to compensate for the difference.