Hysteresis - Calculation method

Hello, everyone:

I have discussed with some colleagues about which is the correct criteria or method to calculate the hysteresis error. I will use as an example a humidity sensor VS output (V) with two (02) up-down cycles. I will attach some figures for better understanding.

CRITERIA 1
1.1: Calculate hysteresis for each input value as an absolute value between each pair up-down for each cycle. I will explain this point later on.
1.2: Calculate the average hysteresis for each input value.
1.3: Maximum average hysteresis is the hysteresis error.

1706811000359.png

CRITERIA 2
2.1: Calculate the average "UP" and the average "DOWN" values for each input.
2.2: Calculate the difference between average values.
2.3: Maximum difference value is the hysteresis error.

1706811149046.png

I do not agree with the CRITERIA 2 as it consideres negative values with positive ones without any absolute values. Therefore, negative errors "cancel" the positive ones.

In fact, by using the CRITERIA 2 with other data within the column "DOWN 2" in such a way it appears the same average "UPs" and "DOWNs", it would obtain hysteresis equals to CERO... it makes no sense because if those cycles are graphed, both show hysteresis error.

1706811649193.png

1706811926450.png

Due to aforementioned discussion, I think the correct method is the CRITERIA 1 as it considers the hysteresis error as an absolute value for each input and each cycle.

I would appreciate your comments in this regard.

Thank you and regards,

Jaime
 
You need to explain how the individual tests are performed, and what standard is used to guage calibration error, before attempting to address hysteresis.
 
If it is only an example, there are countless references to address accurracy and the various subcategories involved. You'll be better off sorting out what works in your case. It won't be simple as there are eighteen categories for static error, and that's just the beginning.

Good luck,
 
Top