CN107402857A - A kind of method of raid card termination of insurance status analysis - Google Patents

Abstract

The invention discloses a method for insurance failure analysis of a raid card, which relates to FA analysis technology and is oriented to mass-produced raid board insurance failure, from insurance failure mechanism analysis, insurance failure factor analysis, insurance failure factor verification, insurance failure prevention and improvement countermeasures Multi-faceted FA analysis, through continuous analysis and verification, directly finds the root of the problem, and completes the insurance failure analysis of the raid card simply and quickly. The invention solves the abnormality of the mass-produced raid card self-check, is beneficial to the standardization of the insurance failure analysis of the raid card, improves the analysis thinking and analysis efficiency of the raid card insurance failure problem, and improves the yield rate of the mass-produced raid card.

Description

A method for insurance failure analysis of raid card

technical field

The invention relates to FA analysis technology, in particular to a method for analyzing the insurance failure of a raid card.

Background technique

A Raid card is a board used to implement the raid function, and is usually composed of a series of components such as an I/O processor, a hard disk controller, a hard disk connector, and a cache. Different raid cards support different raid functions. The Raid card can allow many disk drives to transmit data at the same time, and these disk drives are logically a disk drive, so that the rate of a single disk drive can be several times, dozens of times or even hundreds of times. In addition, raid cards can provide fault tolerance.

Usually in the server, it is implemented by integrating a raid controller or an independent raid card on the motherboard and connecting multiple disks. Raid technology provides higher performance index, data integrity and data availability than usual disk storage, which is of great significance for server data storage. However, the yield rate of mass-produced RAID cards often exceeds the standard, and the problem phenomenon is concentrated. The current analysis method for the insurance failure of the raid card cannot quickly and effectively find the problem point. Therefore, it is necessary to propose a fast FA analysis method for insurance failure of mass-produced raid cards. FA (Failure Analysis) is the reverse thinking and exploration for people to understand the nature of things and the law of development.

Raid card insurance main parameters: 1) Rated voltage: The rated voltage value of the fuse must be greater than the maximum voltage of the disconnected circuit to ensure that the fuse can quickly extinguish the arc (extinguish the arc); 2) Rated current: The rated current value indicates that the fuse is on Current carrying capacity under a series of test conditions; 3) Fuse integral: The fuse integral indicates the energy required for the fuse to blow, in I2t.

Contents of the invention

The invention provides a method for analyzing the insurance failure of a raid card aiming at the needs and deficiencies of the current technical development.

A method for insurance failure analysis of a raid card according to the present invention, the technical solution adopted to solve the above-mentioned technical problems is as follows: the method for analyzing the insurance failure of a raid card is oriented to a mass-produced raid board insurance failure, from the analysis of insurance failure mechanism, insurance Failure factor analysis, insurance failure factor verification, insurance failure prevention and multi-faceted FA analysis of improvement countermeasures, through continuous analysis and verification, directly find the root of the problem, and quickly complete the raid card insurance failure analysis.

Preferably, insurance failure mechanism analysis: mainly analyze the factors of fuse blowing; 1) analyze whether the current exceeds the maximum allowable current of the fuse, if it exceeds, the fuse will blow; 2) analyze whether the intersection point of the current flowing through the fuse and the duration exceeds the current time The safety range of the curve, if it exceeds, the fuse will blow.

Preferably, the insurance failure factor verification includes relevant verification for three failure factors of 1) surge current, 2) short circuit, and 3) plug-in operation.

Preferably, 1) Inrush current: verify whether the inrush current at the moment when the capacitor is powered on exceeds the parameter range of the fuse, and if it exceeds, the insurance will fail.

Preferably, use an oscilloscope to measure the current waveform at the moment of power-on, and judge whether the current waveform satisfies the following four conditions at the same time: a, according to the current value and the current pulse width, the calculated current integral value is less than the insurance nominal fusing integral parameter; b. The actual measured maximum current value is less than the maximum current value allowed by the insurance; c. The rated current value is less than the insurance rated current; d. The working voltage is lower than the insurance rated voltage; if the above four conditions are met at the same time, it indicates that the fuse is working normally .

Preferably, 3) plugging and unplugging operation: verify incorrect and non-standard plugging and unplugging of the board, and judge whether there is a fuse risk according to the current of the insurance of the testing board.

Preferably, use an oscilloscope to measure the current waveform when the board is plugged and unplugged in various situations, compare the current value with the maximum allowable current value of the insurance, and judge whether the surge current exceeds the parameter range of the fuse, and then judge the various types of the board. Whether there is a risk of blowing the fuse during plugging and unplugging operations.

Preferably, when the host computer is turned off, if the board card is inserted and removed correctly, and the oscilloscope does not detect any obvious current changes, then the insurance does not have a fuse risk; when the host computer is not powered off, the board card is plugged in and out, and the current value exceeds The maximum allowable current of the insurance, there is a risk of fuse blowing.

Preferably, when the board card is not installed in place, if the LED light of the board card works normally after power-on, and the current is always below the rated current of the insurance, there is no risk of fuse blowing; if the board card cannot work normally after power-on, If the position of the board is corrected without turning off the power of the host, the maximum current value exceeds the maximum allowable current of the fuse, and the fuse will burn out.

Preferably, when the board is not fully inserted into the PCI slot, the LED lights of the board will work normally after the host is powered on. If the host vibrates or the operator touches the board, a series of current shocks will occur, and the maximum value of the current exceeds the insurance value. The maximum allowable current, the fuse will burn out.

Compared with the prior art, the method for insurance failure analysis of a raid card in the present invention has the beneficial effect that the method is oriented to rapid FA analysis for mass-produced raid board insurance failure, and is beneficial to raid through professional FA analysis. The standardization of card insurance failure analysis has accelerated the analysis of raid card abnormalities, improved the analysis ideas and speed of raid card insurance failure problems, and improved the product quality and mass production yield of raid cards.

detailed description

In order to make the purpose, technical solution and advantages of the present invention clearer, the method for insurance failure analysis of a raid card described in the present invention will be further described in detail below in conjunction with specific embodiments.

In order to effectively improve the yield rate of raid cards, the present invention proposes a method for insurance failure analysis of raid cards, which is oriented towards mass production raid board insurance failure, from insurance failure mechanism analysis, insurance failure factor analysis, insurance failure factor verification, and insurance failure prevention Multi-faceted FA analysis and improvement countermeasures, through continuous analysis and verification, directly find the root of the problem, simply and quickly complete the insurance failure analysis of the raid card, solve the abnormal self-test of the mass-produced raid card, and improve the speed of analysis and the quality of mass production .

Embodiment: 1:

This embodiment proposes a method for insurance failure analysis of a raid card, which includes multi-faceted FA analysis from insurance failure mechanism analysis, insurance failure factor analysis, insurance failure factor verification, insurance failure prevention and improvement countermeasures; through professional FA analysis, improve The analysis idea and analysis speed of the raid card speed up the abnormal analysis progress.

The essence of fuse blowing: current flows through the fuse, causing the fuse to heat up. If the heat cannot be dissipated in time, the fuse will continue to heat up, and the fuse will be disconnected after the temperature reaches the melting point. Factors for a fuse to blow include: 1) Exceeding the maximum allowable current of the fuse, 2) The intersection of the current flowing through the fuse and the duration exceeds the safe range of the current-time curve. For example, if the intersection point of the current flowing through the fuse and the duration is on the right side of the current-time curve, the fuse will blow; then factor 2) of the fuse blowing means that the intersection point of the current flowing through the fuse and the duration is at its current time right side of the curve.

Insurance failure mechanism analysis: mainly analyze the factors of fuse blowing; 1) analyze whether the current exceeds the maximum allowable current of the fuse, and 2) analyze whether the intersection point of the current flowing through the fuse and the duration exceeds the safe range of the current-time curve.

Insurance failure factor verification, including related verification for three failure factors of 1) surge current, 2) short circuit, and 3) plug-in operation.

1) Surge current:

The two ends of the capacitor are close to a short-circuit state at the moment of power-on, and finally a surge current is formed. If the inrush current value of the capacitor at the moment of power-on exceeds the capacity of the fuse, it will cause the insurance to fail.

At the moment of power-on, the capacitance is equivalent to the series connection of an ideal capacitor C and an internal loss resistor Rs, then, the peak current of the capacitor at the moment of power-on is: I=V/Rs.

Main verification: Whether the inrush current at the moment of power-on of the capacitor exceeds the parameter range of the fuse. Use an oscilloscope to measure the current waveform at the moment of power-on, and judge whether the current waveform meets the following four conditions at the same time: a. According to the current value and current pulse width, the calculated current integral value is less than the insurance nominal fusing integral parameter; b. The actual The measured maximum current value is less than the maximum current value allowed by the insurance; c, the rated current value is less than the rated current of the insurance; d, the working voltage is lower than the rated voltage of the insurance. If the current waveform satisfies the above four conditions at the same time, it indicates that the fuse is working normally; otherwise, there is a risk of the fuse blowing.

2) Short circuit: Verify whether the fuse fails due to a short circuit.

3) Insertion and extraction operation: verify incorrect and non-standard insertion and removal of boards, and judge whether there is a risk of fusing according to the current of the test board insurance; the judgment basis is whether the surge current exceeds the parameter range of the fuse.

Use an oscilloscope to measure the current waveform when plugging and unplugging the board under various circumstances, compare the measured current value with the maximum allowable current value of the insurance, and judge whether there is a risk of fuse blowing due to various plugging and unplugging operations of the board.

1) Grasp the correct and standardized current waveform when plugging and unplugging the board: When the host is turned off, plug and unplug the board, and the oscilloscope cannot detect obvious current changes, so there is no fuse risk;

2) Grab the current waveform when plugging and unplugging the board: When the host is not shut down, plug and unplug the board, the current value exceeds the maximum allowable current of the insurance, there is a risk of fuse blowing, and the fuse blowing fails in the actual test;

3) Grab the current waveform when the board is not installed in place: after power-on, the LED lights of the board work normally, and the current is always below the rated current of the insurance, so there is no risk of fuse blowing; after power-on, the board cannot work normally Work, if the position of the board is corrected without turning off the power of the host, the maximum current value exceeds the maximum allowable current of the fuse, and the fuse fuse condition is met, and the fuse will burn out;

The board is not fully inserted into the PCI slot for testing. After the host is powered on, the LED lights of the board work normally. At this time, the host vibrates or the operator touches the board, which causes the board to be plugged in with power, and a series of current shocks occur. If the maximum value of the current exceeds the maximum allowable current of the insurance and meets the conditions for the fuse to blow, the insurance will burn out.

Example 2:

The method for analyzing the insurance failure of the raid card described in this embodiment, on the basis of Embodiment 1, takes the fuse 0603SFF300F/32 as an example to further describe the technical content of the present invention in detail. The main parameters of 0603SFF300F/32: rated voltage: 32Vdc; rated current: 3A; fusing integral: 0.19A2Sec; maximum current: 35A.

FA analysis is carried out from the verification of insurance failure factors, and relevant verification is made for the failure factor of surge current, mainly to verify whether the surge current formed exceeds the parameter range of the fuse when both ends of the capacitor are close to a short circuit state at the moment of power-on. Use an oscilloscope to measure the current waveform at the moment of power-on. Only when the current waveform meets the following four conditions at the same time can it prove that the insurance works normally; otherwise, there is a risk of fuse blown.

The four conditions are as follows: a. According to the current value and current pulse width, the calculated current integral value must be less than the insurance nominal fusing integral parameter (0.19A2Sec); b. The actual measured maximum current value is less than the insurance allowable maximum Current value (35A); c. The rated current value is lower than the rated current of the insurance (3A); d. The working voltage is lower than the rated voltage of the insurance (32V).

1) Measure the power-on surge current waveform under the condition of actual capacitance value: the current value at the moment of power-on is 880mA, the current pulse width is about 10ms, the calculated current integral value is 0.0077, which is less than the insurance nominal value of 0.19, and the current is always If the rated current of the fuse is not more than 3A, the fuse is working normally. Therefore, if the capacitance value is selected properly, there is no risk of insurance failure due to the surge current at startup.

2) Measure the power-on surge current waveform when the capacitance exceeds the upper limit: increase the capacitance by 1000uF, the current value at the moment of power-on is 1.89A, the current pulse width is about 15ms, and the calculated current integral value is 0.0536, which is less than the insurance nominal value 0.19, and the current never exceeds the rated current of the insurance, the insurance is working normally.

3) When the total capacitance increases to 4000uF, the current value at the moment of power-on is 7.5A, and the current pulse width is about 15ms. Since it is not a rectangular wave, the current integral is based on the waveform envelope area, and the current envelope integral value is about 0.4, which is greater than Insurance nominal 0.19, then there is a risk of insurance blown at this time. For the fuse 0603SFF300F/32, based on the above experimental results, when the equivalent capacitance at the rear end of the fuse is greater than 2500uF, the fuse integral value is greater than the nominal value of the fuse, and there is a risk of fuse blown.

FA analysis is carried out from the verification of insurance failure factors, and relevant verification is made for the failure factors of plugging and unplugging operations. Use an oscilloscope to measure the current waveform when plugging and unplugging the board under various circumstances, and compare the current value with the maximum allowable current value of the insurance to determine whether various incorrect plugging and unplugging operations of the board have the risk of blowing the insurance.

1) Grab the current waveform when plugging and unplugging the board: When the host is not shut down, plug and unplug the board, the current is as high as 49.8A, which far exceeds the maximum allowable current of 35A for insurance, and there is a risk of fuse blowing, and in the actual test Fuse blown failed.

2) Grab the current waveform when the board is not installed in place: for example, the board is tilted to the right, and the LED lights of the board work normally after power-on; in the current waveform of the board when it is working stably, the current is always at the rated current of the insurance Below 3A, there is no risk of fuse blowing.

3) Grab the current waveform when the board is not installed in place: for example, the board is tilted to the left. After power-on, the board cannot work normally. If the position of the board is corrected without turning off the power of the host, the fuse will burn out. In this case, in the current waveform when the fuse burns, the maximum value is 143A, which exceeds the maximum allowable current of the fuse by 35A, which meets the fuse blowing condition, and the fuse burns out in actual use.

4) Grab the current waveform when the board is not installed in place: the board is not fully inserted into the PCI slot for testing. After the host is powered on, the LED lights of the board work normally. At this time, the host vibrates or the operator touches the board. Card, forming a live installation of the board, the insurance will burn out. When the host vibrates or touches the board, a series of current shocks occur. The maximum current is 112A, exceeding the maximum allowable current of the insurance by 35A.

From the above analysis results, it can be seen that when the host is not turned off, any plugging, touching, and shaking board operations have the risk of insurance failure; although the actual capacitance value of the board will cause a current shock, the current value is within the insurance range. Within the allowable range, the insurance will not be burned.

The specific embodiment is only a specific case of the present invention, and the scope of patent protection of the present invention includes but is not limited to the above-mentioned specific embodiment, any suitable claims that meet the claims of the present invention and those of ordinary skill in the art. Changes or replacements should fall within the scope of patent protection of the present invention.

Claims (10)

1. A kind of 1. method of raid card termination of insurance status analysis, it is characterised in that towards volume production raid board terminations of insurance status, from guarantor Dangerous Analysis of Failure Mechanism, termination of insurance status factorial analysis, the checking of the termination of insurance status factor, termination of insurance status prevention and improving measure it is multi-party Face FA is analyzed, and by constantly analyzing checking, where the root for directly finding problem, quickly completes raid card termination of insurance status point Analysis.
2. A kind of 2. method of raid card termination of insurance status analysis according to claim 1, it is characterised in that the termination of insurance status machine Reason analysis：Analyze the factor of blown fuse；Including 1）The maximum allowed current whether electric current exceedes fuse is analyzed, if exceeding Then blown fuse；2）The electric current of fuse is flowed through in analysis and whether the intersection point of duration exceeds the safety of current-time curvel Scope, if the blown fuse if.
3. A kind of 3. method of raid card termination of insurance status analysis according to claim 1, it is characterised in that the termination of insurance status because Son checking, including for 1）Surge current, 2）Short circuit, 3）Plug three kinds of Failure Factors of operation do the checking of correlation.
4. 4. a kind of method of raid card termination of insurance status analysis according to claim 3, it is characterised in that described 1）Surge electricity Stream：Verify whether the surge current of electric capacity powered on moment exceeds the parameter area of fuse, if the termination of insurance status if.
5. 5. a kind of method of raid card termination of insurance status analysis according to claim 4, it is characterised in that use oscilloscope measurement Powered on moment current waveform, judge the current waveform whether and meanwhile meet following four condition：A, according to current value and electric current Pulsewidth, the current integration values being calculated, which are less than, insures nominal fusing integral parameter；B, the maximum current value actually measured The maximum current value allowed less than insurance；C, load current value is less than insurance rated current；D, operating voltage is less than the volume of insurance Determine voltage；If meeting aforementioned four condition simultaneously, show fuse normal work；Otherwise fuse melting risk be present.
6. A kind of 6. method of raid card termination of insurance status analysis according to claim 3, it is characterised in that 3) the plug behaviour Make：When verifying incorrect, nonstandard plug board, according to the electric current of test board insurance, fuse melting is judged whether Risk.
7. 7. a kind of method of raid card termination of insurance status analysis according to claim 6, it is characterised in that use oscilloscope measurement Current waveform during board is plugged in the case of various, the current value measured and the maximum allowed current value of insurance are contrasted, sentenced The various plugs operation of disconnected board is with the presence or absence of the risk for causing fuse melting.
8. 8. a kind of method of raid card termination of insurance status analysis according to claim 7, it is characterised in that in the shape of main frame shutdown Under state, correct specification plugs board, then fusing risk is not present in insurance；In the state of main frame does not shut down, board is plugged, Fuse melting risk be present.
9. 9. a kind of method of raid card termination of insurance status analysis according to claim 7, it is characterised in that the board is not installed When in place, if it is upper electricity after board LED normal work, in the absence of fuse melting risk；If board can not be normal after upper electricity Work, the rectification plate card position in the case where not closing host power supply, then there are insurance burn observations.
10. 10. a kind of method of raid card termination of insurance status analysis according to claim 7, it is characterised in that the board is not complete When being fully inserted into PCI slot, on main frame after electricity, the LED normal work of board, if main frame vibration or operator touch plate Card, then there are insurance burn observations.