ASA (VI curve) test on circuit on-line maintenance tester (2)

6. About the Number of Test Channels

The quality of the test depends on the quality of the test signal. The number of test channels mainly affects the test efficiency. The number of channels is required differently depending on the usage requirements. There are three main types:

1. On-line test: At present, there are almost no test clips that can be used with more than 80 pins, so 80 channels basically meet the requirements of use;

2. Circuit board port test: Through the adapter board and the corresponding circuit board socket on the board, lead the test channel to each of the circuit board sockets, and then conduct single/multiple port test. Usually 160 channels can meet most of the usage requirements;

3. Offline test of VLSI: For various VLSI devices, except ASA

Testing, the average user has few other means of testing. The test is similar to a circuit board port test except that the adapter board is a device socket, not a circuit board socket. The number of channels required in this case depends on the number of device feet.

Specifically, only the impact of test signals and number of test channels on ASA testing is discussed here. The general circuit on-line maintenance tester has a variety of functions, other functions will also use these hardware resources. For example, circuit board network test on sink energy test also requires more channels.

7. Increase the detection rate of basic ASA tests

1. Single Port Testing vs Multiport Testing

The so-called single-port test usually refers to all the circuit nodes (device pins) and the "ground" of the circuit board form one port after another, on which the ASA test signal is always added for testing. In single-port mode, a total of N curves are processed for a device with N feet.

The so-called multi-port test refers to the ASA test of any two pins of a device. In a multiport fashion, for a device with N pins, N is processed at most × N curves. Multiport testing is easy to detect faults where the curves of the two pins do not change much, but the curves between the pins change a lot.
On the energy sink tester, there are three different situations to consider for multiport:

A. Multiport symmetry:

When a port composed of two to three feet is processed, it is called symmetrical if a port composed of three to two feet is also processed. This is similar to using a universal meter to test again with a turntable bar. Otherwise, it is called asymmetric.

Because ASA uses positive and negative symmetrical AC test signals, in most cases, the curves measured after signal tuning on the same port are symmetrical with the original curve, that is, no more fault information can be provided. Symmetrical mode is more to facilitate fault analysis and improve efficiency. But there are also asymmetries.

B. Asymmetric with reference foot:

C. Asymmetric without reference foot:

The so-called with reference finger to test 1 foot to 1 foot, 2 feet to 2 feet,... Port made up of N-foot to N-foot. Normally, this port curve must be a short-circuit curve. Otherwise, there may be other problems and all the port curves composed of this foot are not trusted. That is, including a reference foot is helpful to improve test reliability.

2. Sensitivity selection of automatic and manual curves

ASA tests detect faults based on the shape differences between the two curves, but the differences in test data are not proportional to the differences in curves when the data is displayed as a curve. Make an analogy with a mechanical multimeter. The closer the pointer is to the central position, the greater the swing of the pointer caused by the change in the resistance value of the same proportion, that is, the higher the sensitivity of the pointer position to the fault.

For ASA curves, it can be proved that the closer the overall trend of the curve is to 45 degrees, the more sensitive it is to the fault under the voltage-current coordinate (VI curve); If the test curve is a closed graph, such as capacitance, the larger the area enclosed by the curve, the higher the sensitivity to failure.

In order to obtain a higher sensitivity curve, the output resistance of the test signal can be adjusted. The closer the output resistance value is to the equivalent resistance value of the measured node, the higher the curve sensitivity will be.

The sink energy meter allows manual/automatic selection of higher sensitivity curves. Manual Selection Trial Using the feature that test software allows setting test parameters by port, any curve can be modified and retried, which can be judged by the user and selected with higher sensitivity. The disadvantage of manual adjustment is that it is difficult to observe and select curves one by one when they are many, and it is often difficult to intuitively judge which curve has higher sensitivity when its shape is irregular. Automatic adjustment uses an algorithm to automatically select curves with higher sensitivity. Automated selection takes a relatively long time to test.

F. Set test parameters by port, retest

Not all pins of the same device apply to the same test parameters. A typical example is that the output of collector open circuit devices is often high voltage resistant, mostly used to drive digital tubes, relays, and so on, while the input is generally standard level, so the amplitude of test signals for output and input should be different in order to achieve the purpose of comprehensive detection.

The energy sink tester has two setup windows: general setup and special setup. Normal settings are valid for all ports, while a few special ports can be reset in special settings. Special settings have a higher priority than general settings.

After completing the test, if you are not satisfied with the test results, you can also use the single curve processing function to select the unsatisfactory curve with the mouse, open the setup window of the curve, and re-test the curve after resetting.

8. Improving test reliability

In practical use, there are roughly the following reasons that lead to the failure of getting true node curves, which leads to misjudgement of failure. Test reliability refers to the ability of the tester to recognize false ASA curves.

1. Contact problems

Contact problem is one of the common problems that affect fault diagnosis. When the open-circuit curve is tested, it is probably caused by contact problems - oxidation of the pin of the device, non-polished rust-proof coating, wear and tear of the test probe, etc. So in practice, once the open-circuit curve is found, it should be confirmed again - checked, processed, and then retested to avoid false curve. However, when testing hundreds or thousands of curves, it is unrealistic to find the open-circuit curve manually. Therefore, the sink energy tester has the function of automatically judging the open-circuit curve. In all the measured curves, as long as there is an open-circuit curve, the user can give a hint and find it easily according to the hint.

2. Curve stability problems

Due to the reasons of the circuit under test, each measured curve is called curve instability differently under the same test conditions. There are three main reasons for unstable curves:

A. Instability caused by self-excitation

The electrical characteristics of circuit nodes are very different. ASA tests inject a sine wave into the circuit nodes. It is entirely possible for a node to self-excite at exactly this frequency. In fact, this is the reason why ASA tests use sinusoidal wave as the test signal - sinusoidal wave has the least harmonic component, which results in the least possibility of self-excitation.

The phenomenon of self-excitation curve is a large area of scatter and there is no final stable state. Replace the test signal frequency to eliminate it.

B. Instability due to transition

It is caused when the charge and discharge constants of nodes differ greatly. The phenomenon of this curve is that when repeated tests, the curve gradually approaches a stable state and has a final stable state.

C. Instability when testing MOS devices

The first two kinds of instability mainly occur in the circuit board test, and this instability mainly occurs in the test of individual MOS devices. The reason for this instability is that the MOS transistors in the device have different switching processes as the test signal rises or falls.

This instability occurs when the curve jumps up and down, as if two stable curves take turns. The test can be eliminated by grounding both the power foot and the foot of the device.

Unstable curves are prone to misinterpretation. However, when testing hundreds or thousands of curves, it is unrealistic to discover unstable curves manually. Therefore, the sink energy tester has the function of automatically determining whether the curves are stable. In all measured curves, as long as there are unstable curves, it gives prompt information. Users can easily find out where they are, and then process and re-test in the way mentioned above.

9. Improve Human-Computer Interface and Utilization Efficiency

1. Index test results

When hundreds or thousands of curves are processed, fast retrieval can seriously affect efficiency. There are three main retrieval requirements in use:

A. Quickly find abnormal curves, such as open-circuit, unstable, comparatively overshoot curves;

B. Quickly find the curve of a port;

C. Quickly find the port to which a curve belongs.

The sink energy meter has been designed carefully and the following measures have been taken:

A. Page all curves. If it is a single port test, all curves are arranged into one page; If it is a multiport test, the curve of the port with the foot is page 2, the curve of the port with the second foot is page 2, and so on. Support page turning up and down;

B. General and page tips are set. If one of the curves on all pages has an abnormal curve, the total hint will give an error hint. If the error is on the current page, a reminder that the current page has an abnormal curve will be given at the same time, so that the user can first determine if there is a problem curve according to the general instructions, if there is, then use the up and down page flipping function to find the page with the error quickly;

C. Curve arrangement within pages

When learning curves, there are two ways to arrange the curves:

1) First abnormal, then normal arrangement;

Since abnormal curves are the main concern in learning, this method is usually used to arrange the abnormal curves in front for easy handling.

2) Order of pins

Supports three types of curve arrangement when comparing curves

3) In order of error reduction

Comparisons with large errors are automatically placed first. Because overerror curves are the main concern in comparison, this method is usually used to arrange curves with large errors in front for easy processing.

4) In the order in which errors increase

Automatically top results with smaller errors

5) Arrange curves in pin order

D. Index Port by Curve

First, page up and down to find the curve page with the desired pin (there is only one curve page in single port mode). If there are many curves beyond the display area of one screen, you can use the mouse to drag the slider bar to roll the screen to the display area.

E. Curve indexed by port

After a multi-port test on a 40-foot device, how can you quickly find the curve of a port, such as a 20-foot and 5-foot port curve? Indexing curves by port is designed to solve these problems. Open the Port Index window, click 20 with the mouse in the selection bar, page 20 becomes the current page immediately, and click 5 again. Without dragging the scrollbar with the mouse, the port curve will automatically scroll to the appropriate position to display, which is very convenient.

2.Three Curve Coordinates

In most cases, ASA curves are observed in voltage-current coordinates, but for some test results, a different coordinate may be easier to understand and expose the nature of the problem. This is like having a Cartesian coordinate system and a polar coordinate system in geometry. The energy sink meter supports three coordinates: voltage-current coordinate, time-voltage coordinate and voltage-resistance coordinate. Refer to the test examples given in the product description for a more specific understanding.

3. Single Curve Processing

When the desired curve is indexed, double-click the curve with the mouse, and the tester will only retest the curve, the other curves will not be affected. As mentioned earlier, it also allows the curve to be retested after its test parameters have been reset.

Some testers do not support single curve retest. When a curve needs to be retested, all curves must be retested. Low efficiency. Others can not set different test parameters for different curves, which can not fully meet the requirements of use.

10. Other ancillary measures

1. Curve display

Energy sink tester displays curves in point mode - according to actual test data; Line-by-line is also supported -- test points are connected to show curves using lines. The way of point is true, the way of line is striking, users can switch at any time according to their own requirements;

2. Single Curve/Page Display

To see curves in detail, you can switch from page display to single curve display, that is, the entire screen shows only one curve.

3. Show/Hide Coordinate Parameters

When the display coordinate parameters are selected, the amplitude of the sinusoidal wave that tests each curve and the short-circuit current will appear on the axis on which the curve is displayed.

G. Direct/indirect comparative testing

The so-called indirect test refers to extracting the ASA curves of the circuit or device, storing them in the computer, and then testing the same but faulty circuit board or device, showing the curves measured twice at the same time for judging the faults. It can be a well-measured, immediate failure; It is also possible to compare the circuit boards suspected of having faults at any later time. Steps are often called learning or building libraries. Because of this

Retain test criteria automatically in this way - There is no need to always have good results for a good board, so this method is generally recommended.

Direct comparative testing refers to the simultaneous testing of two printed boards or devices and the simultaneous display of them. Some people call it double-board comparison. This way, test results are not preserved.

5. Dynamic/Static Testing

The so-called dynamic test refers to testing a curve repeatedly and displaying test results repeatedly. The tester is always in motion until it is stopped manually, so it is called dynamic test. Some call it automatic patrol inspection. Various unstable curves can be observed by dynamic testing.

Static testing refers to testing only a specified number of times and displaying test results. The curve the user sees is static, so it's called a static test.

6. Find the average curve

This function is often used to build a higher quality standard curve library for a large number of the same objects being tested, such as testing a batch of the same devices. Users can first test multiple devices to get multiple curve files, and then use this function to average multiple curve files into one curve file, which will be used as the test comparison standard in the future.

The technical principles of ASA are not complex, but it is not easy to implement ASA testing into a powerful and efficient testing function. The above is the author's experience and experience in this work for many years. Please correct any inappropriateness.