Home » knowledge » Come and see, design a best front-end for UHF partial discharge online monitoring system~

Come and see, design a best front-end for UHF partial discharge online monitoring system~

According to the IEC 60270 standard, a partial discharge (PD) is a discharge that occurs in a partially insulated area between two conducting electrodes with a gap. Partial discharge is widely regarded as the best early warning indicator of insulation aging of electrical assets within the grid.

According to the IEC 60270 standard, a partial discharge (PD) is a discharge that occurs in a partially insulated area between two conducting electrodes with a gap. Partial discharge is widely regarded as the best early warning indicator of insulation aging of electrical assets within the grid.

When partial discharge occurs, a signal with a wider frequency range is generated, so there are 4 partial discharge detection techniques for different frequency ranges. Ultrasonic detection technology for the 20 kHz to ~200 kHz frequency range, High Frequency Current Transformer (HFCT) detection technology for the 3 MHz to ~30 MHz frequency range, and Transient Earth Voltage (TEV) detection technology for the 3 MHz to ~100 MHz frequency range range, ultra-high frequency (UHF) detection technology for the 300 MHz to ~1500 MHz frequency range. UHF detection technology has high detection sensitivity and is widely used in partial discharge online monitoring systems for gas-insulated switchgear (GIS), transformers and ring main units (RMU).

Partial discharge signal analysis

According to section 7.1 of the Q/GDW11282-2014 standard “Specification for Partial Discharge UHF Coupler Field Detection of Gas-Insulated Metal-Enclosed Switchgear”, a standard PD signal generator can generate the following PD pulse signal characteristics: pulse rise time does not exceed 300 ps, ​​pulse The width is between 10 ns and 500 ns. Then, use this information to build PD emulator signals in Python. The rise time is 300 ps and the fall time is 10 ns. The peak amplitude of the pulse signal is 100 mV, and the peak-to-peak noise is 10 mV. The sampling rate is 10 GSPS and the sampling time is 10 μs. The pulse is placed in the middle of the sampling time, and both the rising and falling waveforms are linearly fitted.

The time domain waveform of the simulated PD signal is shown in Figure 1, and the frequency domain waveform is shown in Figure 2. According to Figure 2, the PD signals with the highest energy are in the frequency range below 1 GHz. With pulse rise times below 300 ps, ​​more energy is distributed over a higher frequency range.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 1. PD signal time-domain waveform.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 2. PD signal frequency domain waveform.

In the modern complex electromagnetic environment, there are many wireless interference signals with operating frequencies between 300 MHz and 1500 MHz between UHF PDs. To eliminate this interference, customers typically choose sub-bands between 300 MHz and 1.5 GHz to capture PD pulses. Under normal circumstances, the GSM wireless communication signal of about 900 MHz will be the largest interference signal. One way to solve this problem is to use a band-reject filter (BRF) to reject signals from 800 MHz to 1000 MHz. A typical sub-band division scheme is shown in Table 1. Of course, the sub-band division is flexible, and customers can adjust it according to the actual electromagnetic environment.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Table 1. Typical UHF PD sub-band allocation scheme

According to the sub-band division in Table 1, we only keep the corresponding energy spectral components of the PD signal spectrum shown in Figure 2, and then perform an Inverse Fast Fourier Transform (IFFT) to study that after the corresponding filtering, the time-domain waveform will be what is it like. The filtered time-domain waveform is shown in Figure 3. According to Figure 3, after filtering, the PD pulse peaks drop. After filtering, the PD pulse rise time is increased and the fall time is decreased. After filtering, of all waveforms, the full band has the largest peak, followed by the band-stop band and the low-pass band. The high pass band has the smallest peak, but still captures the PD pulse.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 3. PD signal time-domain waveform after filtering.

UHF PD Detection RF Front End Using ADI Signal Chain

A UHF PD detection RF front-end board with 4 channels can be developed using the ADI signal chain. A block diagram of one of the channels is shown in Figure 4, and a front view of the entire board is shown in Figure 5.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 4. UHF PD detection RF front-end board block diagram.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 5. Front view of the UHF PD detection RF front-end board.

The first stage of this front end is the RF gain block ADL5611. With a low noise figure (NF) of 2.1 dB and a high P1dB of 21 dBm, the ADL5611 provides high dynamic range. The ADL5611 has a gain of 22 dB and is extremely flat from 300 MHz to 1500 Mhz UHF PD operating frequencies with less than 0.4 dB gain ripple. All these features make the ADL5611 ideal for UHF PD detection applications.

The second stage is an Inductor-capacitor-based 300 MHz to 1500 MHz bandpass filter (BPF) that provides out-of-band interference rejection.

The third stage uses two single-pole four-throw (SP4T) RF switches HMC7992 to implement the band selection circuit. The first RF path is a DC to 800 MHz low-pass path, the second RF path is a 1 GHz high-pass path, the third is a band-stop path from 800 MHz to 1 GHz, and the fourth is a pass-through path. Depending on the RF path selection, customers can select different RF frequency bands to capture PD pulses in the frequency band with no or minimal interference. The HMC7992 features a low insertion loss of 0.6 dB, high isolation of 45 dB, and a high P0.1dB of 33 dBm.

Stage 4 is a 300 MHz to 1500 MHz BPF, which is the same BPF used in Stage 2, which further provides out-of-band interference suppression.

The last stage is the RF logarithmic detector ADL5513, which converts the UHF PD signal into a low frequency signal of tens of MHz. Therefore, an ADC with a sampling rate of 40 MSPS or 65 MSPS can be used to convert the analog PD signal to a digital signal. For PD detection applications, the main required RF detector characteristics are response time and dynamic range. The ADL5513 has a response time as low as 20 ns and dynamic range as high as 80 dB, making it ideal for PD detection applications. The AD8318 RF log detector is also suitable for PD detection applications. Compared to the ADL5513, it has a faster response time but slightly less dynamic range.

Test Results

The board was tested for key performance and Figures 6 through 8 are screenshots.

Figure 6 shows the input port of the ADL5513 from the first stage input to the last stage, setting the S-parameters on the thru path. The graph shows that, from 300 MHz to 1500 Mhz, the gain is about 14 dB, the gain flatness is better than 2 dB, and the input return loss is better than C8 dB.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 6. S-parameters from the first stage input to the last stage ADL5513 input in the pass-through path.

Figure 7 shows the response curve of the output voltage and the power of the input continuous wave signal at the center frequency of 900MHz where the PD works. Two channels were measured using input power. According to the test results, the entire signal chain has a linear response over the input power range of C75 dBm to C5 dBm. The performance consistency between channels is also very good.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 7. Output voltage versus input power.

Figure 8 shows the measured output waveform when a 900 MHz continuous wave signal pulse is input. The signal power is C75 dBm, the pulse width is 5 μs, and the pulse period is 10 μs. According to this waveform, the signal-to-noise ratio of the output signal is still quite impressive when the signal power is as low as C75 dBm.

Come and see, design a best front-end for UHF partial discharge online monitoring system~
Figure 8. Output response for C75 dBm pulsed CW input.

in conclusion

This article shows how to use the ADI signal chain to build a UHF PD detection board. This complete reference design allows the user the flexibility to select different frequency bands to eliminate interference in complex electromagnetic environments. It also meets the requirements of China Q/GDW11059.8-2013 standard.

The Links:   CM15LD-12H 1MBI300F-120