Configure Methods for Processing Raw Flux Signal

Related Products:


Related Documents:





raw flux processing


Describes how to set up the processing methods by: 1) providing generator settings, 2) setting the processing cycle, and 3) configuring the extrema from raw flux signals.

Configure Methods for Processing Raw Flux Signal

Step 1. Enter the Generator Settings

Step 2. Adjust the Flux Cycle

The fields at the top of the gray area identify the cycle.


Complete these steps:

  1. In the Smoothing Length field, enter the number of points for smoothing filter to apply to the raw waveform. This value helps minimize noise while preserving the shape of the flux signal.

  2. In the Cycles to Process field, enter the number of cycles to analyze for the extrema. This number can be no more than 2 less than the Cycles to Acquire on the Acquisition settings tab. See Configure the Flux and Load Channels for more information.

  3. Check the Smooth? box to apply smoothing.

  4. In the HP FFT Index field, enter the FFT index of the raw flux signal used to highpass filter the data to accentuate the extrema “wiggles”.

    This number is a compromise between:

    • Not filtering the meaningful wiggles in the actual extrema

    • Filtering the background cyclic behavior and the first few harmonics of this cyclic behavior

  5. In the Shorted Turn Deviation field, enter the percentage used as the limit above which extrema deviations between the A and B pole are noted as high and potentially as alarm condition.

Step 3. Set the Extrema

re-use_Raw Flux Processing: Extrema Intro

The characteristics of the flux signal can change dramatically across generator loads. In the Processing settings, you can set processing parameters based on the generator load.

The parameters in each row are used for processing the raw flux signal when the generator load is at least the Load value in that row and less than the Load value in the next row. Using the parameter values in the sample below, the parameters in the first row are used between 0% and 30% load, and the parameters in the second row are used between 30% and 67% load.


Complete these steps:

  1. In the Load field, enter the breakpoint at or below which the rest of the parameters in the row are used for processing the raw flux signal. The units are in percentage of full generator load.



    The Load value in last row of the table must be set to 100%.

  2. For 2-pole generators, the RFMS does not require a keyphasor signal to locate the cycle. As part of the initial processing of the raw flux signal, it finds a full cycle by analyzing the waveform itself. The processing searches for regions of the flux signal with large amplitude to establish a threshold level above which the flux signal is considered large enough to have flux wiggles.

    To set how the threshold is determined:

    1. Check the Calc Env. Thresh? box to calculate the threshold based on the Env. Thresh Fraction value. Uncheck the box to use the Env. Thresh value.

    2. If using the Env. Thresh Fraction, enter the value to calculate the threshold.

    3. If using the Env. Thresh value, enter the threshold level.

  3. In the Start Cycle Method column, enter the starting position of the flux signal cycle as determined by one of three methods:

    • Envelope: Based completely on the start index determined by the rough envelope of the HP filtered flux signal.

    • Harmonic XSig: Based on an average of the location where the fundamental harmonic of the raw flux signal crosses the raw flux signal and the location where the harmonic crosses zero.

    • Harmonic XZero: Based on the location where the harmonic crosses zero.

    In practice, the Envelope method works well for flux signals acquired at lower loads while the Harmonic XSig method works best on flux signals acquired at higher loads. The reason for this difference is the transition from a symmetric signal at low loads to a skewed signal at higher loads. So at higher loads, the fundamental harmonic becomes increasingly off-center as the load increases.

  4. At high load, the raw flux signal can exhibit a dip towards zero in the middle of the wiggles, such that the waveform structure is concave with high average values on the edges and lower average values in the middle. With these types of waveforms, the fundamental harmonic of this signal is heavily skewed by this inverted flux signal shape. To assist with the downstream processing used to detect the start of a flux cycle, these portions of the flux signal can be flipped.

    To flip the flux signal:

    1. Check the Flip Flux? box to flip the flux signal.

    2. In the Flip Thresh. Fraction column, enter the option that determines which values of the raw flux waveform are flipped relative to zero. If the value is:

      • Zero: All positive flux values are flipped around the maximum positive value, so that the maximum value becomes zero and zero becomes the maximum. All negative values are similarly flipped relative to the minimum.

      • 0.5: For positive values, only the top half of the largest positive value are flipped for values above zero. For negative values, only the bottom half of the largest negative value are flipped for values below zero.

      • 1: No waveforms are flipped.

  5. Click OK the close the Settings window.

Do you need more help?

Was this article helpful?
0 out of 0 found this helpful