Partial discharge (PD) measurement on high voltage equipment according to IEC 60270 is a globally established procedure for quality assurance. The acceptance levels defined in standards or prescribed by customers are challenges in the test field due to interference. A method for reducing interference, which is mentioned in the IEC 60270, is the differential measurement using PD balance bridges.
The MBB1 together with a single channel PD measurement system, MPD 600 and two CPL542 measurement impedances is a balance bridge measurement system which can reduce noise. Noise suppression can be achieved because of superimposition of the external interference in both branches
PD signals and common mode disturbance signals show differences in the polarity in the two branches of a PD measurement circle. If PD is not measured with a single measuring impedance but rather as a difference signal according to Figure 1, common mode (disturbance) signals (green) will be reduced whereas PD signals (red) will be superimposed constructively. As a consequence, the SNR of the measurement will increase. This principle is used since the early days of PD measurement and is also stated in the IEC60270.
Measurements like this require symmetry in terms of impedance (capacitance, tan δ, inductance) between Device Under Test (DUT) 1 and DUT2/CK. As a truly symmetrical bridge-type setup is hardly ever given under real conditions, the signals have to be balanced to get better noise reduction. In classical measurement bridges this is done by adjusting the measuring impedances (see Figure 2).
In the automatic mode, a calibrator is connected as common mode source (so between HV and GND) and the software calculates the parameterization according to the measured data.
Figure 3 shows a setup with a 6.6kV voltage transformer (VT) as test object and a 1nF coupling capacitor. The DUT has a capacity of
350 pF and a known insulation defect that leads to internal PD above 5 kV. On the top electrode of the test transformer, a wrench was installed to induce corona discharges. The PD signals were decoupled in the ground path of Ck (CD1) and DUT (CD2).
The synchronization voltage was measured on CD1, and the PD signals were measured within a frequency range of 250 kHz ± 150 kHz. During the adjustment and calibration, the weighting ratio was detected to be 1:2.45. To assess the noise suppression, common- mode pulses with a charge of 2 nC were injected after calibration
Figure 4 and Figure 5 show the PRPD pattern at a test voltage of 5.5 kV. All patterns cover a time span of 20 s. The wrench causes strong corona discharges on CD1 and CD2. In Balance Bridge Mode, the disturbance is reduced and cannot be detected anymore. The insulation defect causes strong internal PD activity with similar PRPD patterns.
For technical enquiries, please contact Seokhoon Hong (Regional Application Specialist – Partial Discharge) via email firstname.lastname@example.org or ENTEC A&T’s Email: email@example.com