In the daily work of FAE, we often receive feedback from users that although the system is already running normally, the results obtained when testing the quantified waveform data are not satisfactory. Why is there such a discrepancy? The problem may lie in the use of testing methods and equipment!
Before conducting a comprehensive investigation into whether it is a system product issue (with a huge workload, time-consuming and labor-intensive), we must first check the testing environment, methods, and methods to see if they have been "tested correctly"? How to choose the correct testing method is particularly important!
Ⅰ.How to select a suitable oscilloscope based on its key indicators
As a commonly used high-precision testing instrument, the oscilloscope can transform invisible electrical signals into visible
images, making it easy for people to study the changing process of various electrical phenomena. The correct use of an oscilloscope is crucial, as testers often experience unnecessary trouble due to incorrect parameter settings that result in "measured data" being significantly different from the actual working state of the system.
The three key indicators of an oscilloscope are bandwidth, sampling rate, and storage depth.
1. Bandwidth: refers to the frequency range when the response causes the output amplitude to decrease to 70.7% (-3dB).
With the continuous development of high-frequency power switch and rectifier technology, and the continuous improvement of power supply operating frequency, the current power switches on the market, such as GaN MOSFETs, SiC MOSFETs, and SiC Schottky rectifier tubes, have an on/off time of less than 5ns (with an off frequency exceeding 200MHz). In the engineering measurement process, to observe such rapidly changing signals, a measurement system with sufficient bandwidth is needed, which is not only the bandwidth of the oscilloscope, The bandwidth of the probe should also be sufficient.
The bandwidth of commonly used differential probes and oscilloscopes is 100MHz, which can meet the needs of daily testing.
The higher the bandwidth, the wider the range of higher-order harmonics of the measured signal that can be collected, and the less distortion of the measured signal. However, the bandwidth of the probe is not necessarily better. The higher the bandwidth, the more frequencies are introduced, and the more noise signals enter. Taking the ripple noise testing oscilloscope as an example, the 20MHz bandwidth limit needs to be enabled for measurement, which limits the bandwidth. Similarly, when there is too much noise interference in the low-frequency signal being tested, Bandwidth limitation can also be enabled on differential probes (5MHz) or oscilloscopes.
2. Sampling rate: Refers to the number of data points that can be collected per second. Generally speaking, the sampling rate indicator of an oscilloscope refers to the highest sampling rate that can be achieved during operation. Storage depth=sampling rate × Sampling time. When an oscilloscope displays a waveform on the screen, it refers to the number of
waveform data. The waveform displayed on the oscilloscope screen is composed of many sampling points, and the number of all sampling points is the storage depth.
What is the impact of storage depth on measurement? We add a square wave with a frequency of 1KHz and an amplitude of 2V to the oscilloscope, and use a 28M storage depth oscilloscope to intercept a 14S signal. At this time, the sampling rate is 2Msa/S and the amplification is 2000 times, but it is still a square wave.
When using a 28K storage depth oscilloscope to intercept a 14S signal, the sampling rate is 2Ksa/S, and the amplification is 2000 times, the resulting waveform is distorted.
From this example, it can be concluded that with the same sampling time, the larger the sampling rate, the deeper the storage depth of the oscilloscope, and more details can be seen in the saved waveform. During testing, ensure that your sampling rate is sufficient to avoid waveform distortion caused by long sampling times. The maximum acquisition rate of a general oscilloscope can reach 4MSA/s in rolling mode, and even higher in triggering mode.
Taking the stress debugging waveform of high-power housing LMF1000-20Bxx product as an example:
When developing, debugging, and testing products, a 4GSa/s high-precision acquisition four channel oscilloscope is usually used, which truly displays the high-frequency signals and transient working data of the product, and can comprehensively evaluate the reliability of the design through data.
Ⅱ. Precautions for using oscilloscopes
1. The oscilloscope must be calibrated when connected to a new passive probe or inserted or unplugged probe for use, otherwise the test results may not be accurate (ripple test results with an error of more than 10mV). During measurement, the probe ground wire should be kept as short as possible. The compensation steps for the probe are as follows:
Connect the probe to a vertical channel, and then connect the probe tip to the square wave reference signal of the oscilloscope;
Observe the square wave reference signal and adjust the compensation capacitance. The adjustment method can be seen in
the following figure;
2. The oscilloscope and probe need to be impedance matched. A general oscilloscope has a switchable matching resistor of 1M Ω (general circuit) and 50 Ω (high-speed circuit) at the input end, which is correctly matched with the probe to reduce the load effect impact of the tested circuit.
3. When grounding the power cord of the oscilloscope, it is necessary to avoid using a regular probe to directly connect to products powered by the power system. Please use a differential probe for testing or use an isolation transformer to power the oscilloscope, or use a floating ground measurement (without a ground wire power cord connecting the oscilloscope) to avoid ground wire noise interference from the real data (the negative terminal of the passive probe is connected to the
power PE of the oscilloscope). For specific comparison, please refer to the following figure:
4. Do not use passive probes for EMC testing, all differential measurements should be used to prevent PE surges from introducing surge signals to the oscilloscope when the oscilloscope is grounded, resulting in damage to the oscilloscope or power loss of the tested product output (abnormal test results). The power supply line of the surge tester and the oscilloscope power supply should be connected
o mains power.
