Range Timers vs. Rate of Change

When rate of change calculations are performed on coffee temperature automatically during a roast, the result is a data series that can be very helpful. Changes in this data series are obvious more quickly than in the raw temperature data and with practice it's possible to make the control adjustments needed to stay on the roasting plan with better precision.

While this data series is very useful in helping match roasting plans with production batches on the same machine or at machines with appropriate calibration adjustments applied, the data is inherently non-portable in other cases. The most common mistake right now is attempting to read too much into trends in this data over small amounts of time. For example, you might decide that you want the rate of change to consistently decline throughout the roast. Looking at the average rate of change over key ranges during a roast, it's reasonable to decide that such a plan is a good starting point when working with a new coffee (note there are several reasonable starting points for roasting plans depending on your machine and preferred roasting style). If you look at example graphs showing rate of change from differently sized thermocouples or different probe placements, or if you look at similar graphs showing raw measurement data against measurements calibrated to a different roaster, you'll usually see that while these large scale trends are comparable, looking at this over smaller time ranges such as minute to minute changes, you'll find several regions in which one set of measurements produces a descending rate of change and another simultaneously produces an ascending rate of change. It is not possible to determine which of these is more valid. The important point is that they should be consistent from batch to batch.

Range timers in Typica provide a way to capture the duration of significant temperature ranges during a roast. These raw duration values when tied to easily observable physical changes in the coffee are considerably more portable in describing a roasting plan than temperature and rate of change data. It is also possible to specify desired durations in a roasting specification in Typica for comparison with reported values at the end of the batch. Range timers are less useful for making precise control adjustments during a roast, but more useful as one of several possible metrics for production consistency.

Both of these features are useful, but it's important to understand the difference so you can use each of these appropriately.