The voltage applied to a flickermeter is processed in five blocks, as specified in [1]. The NPL implementation of the first four of these blocks is given in Design of NPL Flickermeter. The purpose of the fifth block is to sort the flicker readings obtained from the first four blocks into bins, such that a Cumulative Probability Function (CPF) can be obtained. This document describes the method used for this in the NPL Reference Flickermeter, which is used in the NPL Flickermeter Calibration Service.

For a complete description of the NPL Flickermeter see Design of NPL Flickermeter and to find out how this is used to calibrate commercial flickermeters see NPL Flickermeter Calibration Service.

Description of the Classification Routine

The flicker level output from block 4 is divided into 200 000 logarithmically spaced classes, using the following equation,

(1)  

where,

C is the class number and is equal to  ,

S is the scaling factor which must be multiplied by the output of block 4 to give a maximum output of 1 for 8.8 Hz sinusoidal modulation with a modulation depth of 0.25 % (See Design of NPL Flickermeter for more information),

F is the perceptibility output of block 4 before multiplying by the scaling factor,

C_M is the maximum class number =100 000, i.e. half the number of classes,

INT indicates taking the lowest integer value of the parameter in brackets,

C' is the offset class number, offset by C_M.

The maximum class number, C_M, is added to the class number, C, in equation (1) in order to prevent negative class numbers resulting from the classification process, so that the indices of the array holding the classified data are positive.  This results in a range of values for F of  to 1 that can be processed by the flickermeter model, as any values for F less than  would result in a negative offset class number, and therefore a negative array index.

A cumulative probability distribution is then calculated based on the number of samples to fall into each class.  The percentiles defined in Section 4.10.1 of IEC 61000‑4‑15 are calculated by searching for the class number of the flicker level exceeded for the relevant percentage of the test time.  Linear interpolation is used between classes where the percentile value does not fall exactly into a particular class.  The eventual percentile value, P_X, is then calculated using the following equation.

(2)   

where,

(3)    ,

X is the percentile index = 0.1, 0.7, 1, 1.5, 2.2, 3, 4, 6, 8, 10, 13, 17, 30, 50 and 80 and is the searched for class number of the given percentile, after interpolation.

The classification routine results in a maximum error of 130 ppm in the percentile values.

References

[1]    IEC 61000–4–15, Electromagnetic Compatibility (EMC) – Testing and measurement techniques – Flickermeter – Functional and design specifications, Published by The International Electrotechnical Commission, Amendment 1, 2003.