【PM8000】Duration of the ITI curve

[Miyai]

Hello,

What is the basis for the duration used for the PM8000's ITI curve?

I checked the waveform and the duration of sag was 16msec.

However, in the ITI curve, it is described in 30 msec.

Frequency: 50Hz

Waveform  logging : 32x54

I'm guessing:

・The judgment of sag is calculated by the RMS value every half cycle.

・The RMS calculation is performed with 256 sampling points.

　(It is not calculated at the 32 points set by waveform.)

Is this understanding correct?

Judging from sampling 32 points, sag returns in one cycle and does not match the result of the ITI curve. Therefore, I would like to know the details.

Best regards,

Miyai

[Tom_Stevenson]

Hi Miyai - can you post the details of the sag from the event log

Thanks

Tom

[Charles_Murison]

Hello @Miyai ,

Correct the Sag/swell duration is calculated based on the one cycle RMS data calculated every half cycle and not the waveform data itself. Given just a waveform it is not possible to know exactly when the RMS cycle "began" but can calculate the RMS values for each cycle to estimate possible times. Note that the sag/swell module will see single RMS values every half cycle and not a continuous stream like one can calculate with the waveform. 

Regards,

Charles

[sesa56307_bridg · Accepted solution]

To add to the response from Charles, the reason the meters calculate the sag magnitude and duration this way is because that is the standardized definition in the IEC 61000-4-30 standard, with which the PM8000 fully complies. The standard also requires that these "one cycle every half cycle" RMS values are aligned with the fundamental zero crossing on each phase, for added consistency.

Looking at the waveform, it does appear to me that there is at least one phase where there were three consecutive "one cycle every half cycle" windows where the RMS value would be impacted. Three half-cycles means 1.5 cycles, which is precisely 30ms for a 50 Hz system. So the 30ms duration in the ITI curve seems expected to me, given the standardized detection method.

[Miyai]

Hi  @sesa56307_bridg , @Charles_Murison , @Tom_Stevenson 

Thank you for advice.

I would like to confirm my understanding, Is it correct that the sag calculation is the flow of calculation 1 -> calculation 2 -> calculation 3 in the figure?

If  above is correct, I could see from the waveform that the sag occurred for 1.5 cycles.

Best regards,

Miyai

[sesa56307_bridg · Accepted solution]

Hi @Miyai,

Yes, your understanding is correct. The IEC 61000-4-30 standard calls the sequence of RMS values from those windows the "U_rms(1/2)", or the "r.m.s. voltage refreshed each half-cycle". And it requires that "the duration of a voltage dip is the time difference between the time when the U_rms(1/2) falls under the threshold and the time when U_rms(1/2) passes above the threshold plus hysteresis."

Based on that theory, I agree that the sag duration of 1.5 cycles seems consistent with the waveform you showed.

Best regards,

David Tuckey

[Miyai]

Hi　@sesa56307_bridg 

Thank you  for support.

I could understand.

Best regards,

Miyai