The transformer turn ratio tester is mainly used to measure the voltage ratio or current ratio of power transformers, special transformers, and transformers. Its core function is to verify whether the winding turns match the design requirements through high-precision electrical parameter acquisition and algorithm processing. Sensitivity, as a key indicator for measuring the performance of the device, directly affects the accuracy and reliability of the detection results. This article systematically analyzes the core elements that affect sensitivity from four dimensions: hardware architecture, signal processing, environmental adaptability, and operational standards.
1、 Hardware Design and Component Selection
1. Sensor accuracy level
Hall effect sensor: used for non-contact current detection, sensitivity depends on magnetic flux density resolution (typical value ± 0.1% FS). Industrial grade chips with linearity better than ± 0.05% and temperature drift less than ± 50ppm/℃ should be selected.
Precision sampling resistor: Using a manganese copper alloy shunt with low temperature drift (≤± 5ppm/℃) and low parasitic inductance, the stability of the resistance directly affects the signal-to-noise ratio of microampere current measurement.
ADC conversion module: The quantization noise of a 24 bit ∑ Δ type analog-to-digital converter determines the minimum distinguishable voltage difference, and ideally, the signal-to-noise ratio (SNR) should be ≥ 100dB.
2. Characteristics of incentive sources
Sine wave generator: The output waveform distortion (THD) should be controlled at<0.3%, and the frequency stability should reach ± 0.01Hz to avoid harmonic components interfering with the fundamental phase detection.
Power amplifier: Dynamic load adjustment rate<± 0.5%, ensuring constant output amplitude under different capacitive/inductive loads.
3. Shielding and grounding design
Multi layer PCB layout: The simulation front-end adopts a four layer board structure, with the power layer tightly coupled to the ground plane to reduce radiation interference caused by circuit area.
Faraday cage protection: The grounding resistance of the metal shell of the whole machine is less than 2m Ω, effectively suppressing the coupling of power frequency electric field (50/60Hz) and high-frequency noise (>1MHz).
2、 Optimization of signal processing algorithms
1. Digital filtering strategy
Adaptive notch filtering: For power grid harmonics (3rd~19th order), an IIR filter is used to track and attenuate specific frequency interference in real time, with a Q factor set to 8-12.
Wavelet threshold denoising: perform multi-scale decomposition on noisy signals, preserve abrupt features while eliminating white noise, and improve peak signal-to-noise ratio (PSNR) by about 15dB.
2. Synchronous sampling technology
GPS timing synchronization: using satellite clocks to achieve timestamp alignment of multi-channel data, with a time deviation of<± 1 μ s, ensuring the accuracy of phase difference measurement.
PLL frequency doubling: Multiply the input signal to fs=N × fin (N=1024), so that the number of sampling points in each cycle is ≥ 512, meeting the requirements of FFT analysis.
3. Compensation algorithm model
No load loss correction: Establish a mathematical model of iron loss Pa+copper loss Pcu=k · U ²+R · I ², and automatically deduct additional errors caused by excitation current through preset parameters.
Leakage suppression algorithm: Based on Smith predictor, a joint simulation model of inverter transformer is constructed to dynamically compensate for phase shift caused by leakage inductance.
3、 Standardization of operating procedures
1. Wiring specifications
Four wire measurement method: a dedicated testing clamp is used to clamp the tested winding terminals to avoid additional voltage drop caused by wire resistance.
Grounding consistency: All devices are grounded together and the grounding resistance is less than 4 Ω to prevent the formation of circulating currents due to ground potential differences.
2. Principle of Range Selection
Match according to the rated capacity of the transformer to be tested:
S ≤ 10kVA → Optional 0.1%~1% full-scale gear
10kVA
S> 100kVA → 1%~10% gear
The response time for automatic range switching is less than 50ms to prevent overload impact.
3. Preheating and Calibration
Power on self-test: Run the built-in diagnostic program to check for internal reference source (REF) voltage fluctuations of<± 0.01%.
Periodic calibration: Use standard transformers (PT/CT) for traceability calibration every quarter, with an expanded uncertainty U=k=2 ≤ 0.05%.
4、 Trends in cutting-edge technology development
1. AI assisted diagnostic system
Integrating Convolutional Neural Networks (CNN) for deep learning of historical data, establishing a fault feature library, and achieving early warning of inter turn short circuits.
The lightweight YOLOv5 model is deployed in edge computing units, and the local reasoning speed is less than 200ms.
2. Breakthrough in quantum sensing
The superconducting quantum interferometer (SQUID) is applied for weak magnetic field detection, and theoretically can advance the sensitivity to the fT/√ Hz level.
The room temperature diamond NV color center magnetometer has entered the laboratory stage and is expected to replace traditional Hall elements.
3. Photon counting technology
Single photon avalanche diode (SPAD) array is used to achieve flying safety level current detection, and when combined with optical chopper, it can achieve absolute measurement.
The femtosecond laser pumping detection scheme is exploring the observation of sub picosecond transient processes.