Application of RTO Oscilloscope Mask Test Function in Capturing Occasional Errors

Whether there is an abnormal signal or not has always been a question. But when our products encounter occasional faults, we will make bold assumptions and suspect this and that. Next, we wish the oscilloscope could continuously collect signals for half an hour or an hour without missing a single one. But unfortunately, the oscilloscope has never been able to do this job. Every time the oscilloscope collects a screen of signals, it will "pause" for a while before proceeding to the next time. It keeps collecting and stopping. How can we expect the oscilloscope to "catch" abnormal signals? The method introduced in this article should be worth learning from.

In daily circuit testing, it is often difficult to capture and measure occasional errors in the circuit due to various reasons. The difficulties usually lie in:

• Occasional errors have a low probability of occurrence. The oscilloscope cannot capture abnormal signals due to a certain dead time.
• The time and amplitude of occasional errors are uncertain, and it is unclear how to set the trigger conditions for capture.
• How to measure occasional error signals accurately and quickly.

For the first and second difficulties, the oscilloscope needs to have a high waveform capture rate and a more flexible triggering method to capture the discovered signals. For the third difficulty, the oscilloscope needs to have a good operating interface and fast and accurate measurement functions.

This article uses the R&S RTO1044 oscilloscope to test occasional interference signals in circuits as an example to briefly summarize the methods of capturing and interfering with occasional signals.

After the circuit under test is powered on, press Autoset, set the trigger mode to rising edge trigger, and set the mode to Auto. The waveform displayed on the oscilloscope is as shown in Figure 1. It can be seen that it is a standard AC waveform of 220V/50Hz (using a high-voltage probe with 100:1 attenuation). The waveform is displayed stably, and no interference signal is observed.

Figure 1 Oscilloscope waveform after the circuit is powered on

In order to accurately observe whether there is an interference signal, turn on the afterglow function of the oscilloscope, and the waveform display is shown in Figure 2. The afterglow function cumulatively displays all the historical waveforms captured by the oscilloscope, and the depth of the waveform brightness represents the number of waveforms that appear. It can be clearly seen that a burr interference signal appears near the trough of the sine wave.

Figure 2 Oscilloscope afterglow function display waveform

In this process, many engineers will ignore another important indicator of the oscilloscope: waveform capture rate. As the name implies, waveform capture rate refers to the number of waveforms captured by the oscilloscope per second and displayed on the screen. The signal processing process of a modern digital oscilloscope is shown in Figure 3. The waveform of the oscilloscope is constantly changing, but in fact, between two waveforms displayed on the screen, a large number of signals that meet the trigger conditions are "missed", that is, they are not displayed on the screen. Data acquisition process and display processing process. The data acquisition process is implemented through the ADC chip and is very fast. The bottleneck lies in the process of "background data processing -> screen display". The R&S RTO oscilloscope has a waveform capture rate of 1 million per second, which greatly shortens the display processing time of the oscilloscope and greatly reduces the "missed" waveforms.

Figure 3 Signal processing process of modern digital oscilloscope

After observing the glitch signal in the circuit under test, we hope to capture it stably for observation and measurement. Through the single-step triggering "RunSingle" method and the waveform shown in Figure 1, we can see that the glitch signal is occasional and is not always "accompanied" by the sine wave. Of course, the "RunSingle" method can still be used, but for abnormal interference with a very low probability of occasional occurrence, it will greatly waste time and be inefficient. In fact, you can use the mask testing function (i.e. Mask Testing) of RTO to manually draw a template area (i.e. violation area) on the screen. When the waveform touches this area, you can stop the oscilloscope from running, i.e. "Stop", quickly capture the abnormal signal, measure, save the waveform, and perform a series of operations. Then, for the glitch position and amplitude observed by the afterglow function in Figure 3, you can simply draw a violation area at the glitch position to capture it. However, for more common situations, such as when the circuit runs for a long time and the form of interference (i.e. time, amplitude, width) is uncertain, manually drawing the violation area is really a headache. In this case, you can use the "template setting based on waveform reference" function of the RTO oscilloscope. The setting method is shown in Figure 4. Select "Waveform" in Mask Definition, select the channel of the reference waveform in Source, and click "Copy to". At this time, you can see on the screen that all areas other than the normal signal sine wave are set as violation areas. Considering the certain degree of fluctuation caused by the normal signal affected by noise, adjust the two parameters of "Horizontal Width" and "Vertical Width" to reserve a small amount of "room".

Figure 4 Template setting method based on waveform

After setting OK, as shown in Figure 5, you can see that the surrounding area of ​​the sine wave signal has been set as the violation area. Then, once there is any glitch interference signal, no matter where it appears, as long as it causes the original sine wave signal to be distorted, it can be captured.

Figure 5 Template using waveform as reference

Then, zoom in on the interference signal and turn on cursor measurement. As shown in Figure 6, you can see that the negative pulse width of the interference pulse is about 200us. In fact, the test guide document for this project states that the negative pulse width of the interference pulse is about 10us. The problem arises: the interference signal is captured, but the pulse width measurement does not match the guide. Why?

Figure 6 Pulse cursor measurement

In fact, if you observe carefully, the reason is quite simple, "always be vigilant about the sampling rate". In general, the oscilloscope has a fixed storage depth (1K) by default. After pressing Autoset, the sine wave of the measured signal has a large period (50Hz). When adjusting the time base, as the sampling time increases, the oscilloscope will reduce the sampling rate to only 10KHz. The 10KHz sampling rate is no problem for restoring a 50Hz sine wave without distortion, but it is obviously not enough for narrow pulse signals containing high-frequency components, thus causing distortion and pulse width widening.

Therefore, the recommended operation method is to first fix the sampling rate to ensure that the sampling rate can capture at least 5 points on the rising edge of the signal, and then adjust the time base to make full use of the oscilloscope's storage depth. The setting method is shown in Figure 7. Adjust the sampling rate to 500M/s, and then adjust the capture signal time length to about 100ms.

Figure 7 Fixed sampling rate setting

At this time, the burr part is magnified for measurement. As shown in Figure 8, it can be seen that the negative pulse width of the burr is only about 11us, which is consistent with the test instructions.

Figure 8: Enlarging the burr after fixing the sampling rate

So far, we have accurately understood the characteristics of the interference signal in the circuit. The negative pulse width is about 11us, and the positive pulse width is slightly larger. RTO1044 can capture signals with a minimum pulse width of 50ps. Therefore, in order to stably capture interference signals, you can set the trigger mode of the oscilloscope to width (pulse width), as shown in Figure 9. Set it to positive pulse width trigger, set the width to 14us±5us, and change the trigger mode to Normal to capture only signals with interference. Since the horizontal time base is relatively large, the width of the glitch signal is very small. In order to avoid the grid lines of the screen blocking the glitch signal, the grid lines can be removed, as shown in Figure 10.

Figure 9 Pulse width trigger setting

Figure 10 Screen display grid line settings

As can be seen in Figure 11, the oscilloscope can stably capture interference signals, and the trigger point is stably maintained at the end of the positive pulse interference. The RTO1044 oscilloscope also provides a rich parameter fast measurement function, which can test up to 100 parameter values ​​at the same time. Turn on the Measure function and set the measurement of the negative pulse width (Neg Pulse), then the oscilloscope can automatically measure the parameter value. The current measurement is 10.008us, which is fast and accurate compared to the cursor measurement method.

Figure 11 Capturing interference signals based on pulse width

Summarize:

• The waveform capture rate of the RTO oscilloscope can reach up to 1 million/second, so it is very helpful for capturing occasional abnormal signals;
• Rich measurement functions can measure parameters quickly and accurately;
• Always be vigilant about the sampling rate to avoid waveform distortion due to too low a sampling rate, which ultimately leads to incorrect measurement results;
basic standard: there should be at least 5 sampling points on the rising edge of the measured signal;
• Utilize the template test function of the RTO oscilloscope to quickly capture occasional abnormal signals. The operation ideas are as follows:
turn on the afterglow to view the abnormal signal -> set the template and define the violation area -> observe the abnormal signal characteristics -> set the trigger method and trigger mode to stably capture occasional abnormal signals -> perform parameter measurements as needed.