Noise/Inconsistency In RPM Sensing

[AdamBlueR]

The RPM sensing is prone to noise and errors in the reading. Occasionally, a reading wildly out of realistic bounds is logged. This takes a lot of effort to remove from the CSV.

Perhaps the RPM sensing sketch on the Arduino can be improved?

Otherwise, perhaps constraining the inputs and adding error rejection could help. For example, having a parameter for believable RPM (like 10,000), or an algorithm to detect implausible high spikes. In that case, the RPM could be recorded as null for that point, or the previous one copied?