1, BIOS chip failure or BIOS damage by virus.

After the host is powered on, the BIOS code program should be added first. First, the system reads the first instruction in the BIOS into memory and explains its execution. This instruction is the same for both AWARD and AMI, and it is a 3-byte jump instruction. The next step is to read the BIOS boot block to complete the detection and loading of the BIOS itself. After the system reads the BIOS code into the memory, the CPU will start the self-check and startup process according to the code program. There are not many BIOS failures. Occasionally, some storage units of BIOS chip will fail due to aging, resulting in incomplete program code, failure to complete system function call and failure to start normally. There are also viruses such as CIH that maliciously rewrite the BIOS code of the motherboard, causing the host to fail to start. After startup, the host can be powered on normally and the fan works, but the self-check process cannot be completed. At the same time, the display has text display and the host has alarm sound. Of course, it may be serious that only power can be supplied, the display has no image, and the host has no alarm sound.

2, graphics card failure

When the graphics card BIOS is damaged or the AGP communication circuit fails, the initialization of the graphics card cannot be completed, which causes the host to stop at the graphics card BIOS for a long time and crash. Note: Motherboards that have been used for two or three years, especially machines used in Internet cafes, have lost the capacity of the filter capacitors around the CPU due to long-term high-temperature operation, which leads to the increase of AC components in the CPU power supply, indicating that it is not easy to start the machine, and it needs to be turned on several times before it can normally start into the desktop. In the process of self-inspection, it often crashes and the fault performance is unstable and changeable. 3, BIOS self-check

The system will not start the BIOS code check until the video card self-test is completed. The BIOS version number and motherboard model are usually displayed on the monitor. 4. L2 or L 1CACHE of CPU is not completely damaged or other circuits of CPU are aging.

When the host completes the initialization of the graphics card, it begins to detect the CPU self-check. When the L/Kloc-0 and/or L2 of CPU fails, it will stay in "check NVRAM" for a long time. When this fault occurs, some machines can still enter CMOS settings. We can manually turn off L 1 and L2 of the CPU, and then try to see if the machine can pass the self-test. If possible, the CPU can basically be used at a reduced speed. If not, you can only replace the CPU to solve the problem. Most cases of partial CPU damage often show that the system reports BIOS error (such as CIH virus damage), L 1 or L2 error, and fails the self-check, so it can only enter the safe mode and automatically restart when it is started.

5. Memory failure

After the system completes the CPU test, the memory test stage is carried out. If we turn off the POST quick test option, we will not see the memory test interface for a long time (three times in a row) when we start the computer, and only test it once.

Memory failure or poor contact with memory due to memory deformation, memory finger oxidation, memory metal reed deformation or breakage, and low or high memory power supply voltage will all lead to memory self-test failure and crash. 6. PNP device initialization 7. The system can't complete the IDE device detection 1), the interfaces at both ends of the hard disk data cable are in poor contact with the hard disk or motherboard, the plug-in is not real, or the hard disk cable is of poor quality, which leads to the disconnection of some ground wires.

2) The performance of hard disk and CD-ROM drive is partially damaged, which makes the host unable to complete the self-check process and start.

3) If the CD-ROM drive cable is connected reversely, the system self-check time will become longer, and even the CMOS setting will crash.

4) The master-slave disk capacity limit jumper of the optical drive or hard disk is set incorrectly, and the system cannot pass the self-check.

5) For example, the motherboard of Huaqing M266A must use an 80-pin hard disk cable, and the hard disk must be set on IDE 1. If we set a hard disk on IDE2, we will call the police during the self-check and the system will hang up.

6) Generally, optical drives are connected with IDE2 interface, but motherboards with new functions will make suggestions on the normal configuration of the machine during self-inspection. For example, when MSI's MS-6368 motherboard is self-checking, if we don't use DMA66 data cable, the prompt as shown in the following figure will appear, but the system can still be started normally. Note: If the ESCD data is not updated, it will not affect the system self-test. The newly added settings will not be recognized by the system during the self-inspection, and there is nothing unusual in the self-inspection. 8. Wrong floppy drive connection or wrong floppy drive setting in 8.CMOS.

For the failure caused by the reverse connection of the floppy drive line, the floppy drive light will always be on, but it will not affect the normal startup of the system.

If the software AB drive in CMOS is set incorrectly, or the floppy drive cable is loose or wrong, a prompt of "press F 1 to continue" will appear, which will cause the system self-test to be suspended. If we don't want to configure or use floppy drive, just press the "F 1" key, and the system will continue to start and complete the system loading process. 9. When the CMOS battery is dead, the configuration information is lost or the CMOS data is set incorrectly, the computer will have a "CMOS checksum error" during the self-check, and the words "press F 1 to continue" need to be prompted. This kind of problem is generally not fatal. As long as you press the "F 1" key, the computer can continue to start and complete the system loading process. But for ordinary computer users, if the computer does not enter the desktop normally after booting, but stops, it will be mistaken for the computer is broken. Solution:

1), there is no CMOS battery, just replace the CMOS battery directly, and a +3V lithium battery motherboard can last for about 3 years. Sometimes CMOS information can't be saved or lost, not because the CMOS battery is dead, but because there is something wrong with the related circuit.

2) If the checksum error still occurs when the CMOS battery voltage is not lower than +2.8V, we need to press "DEL" or "F2" to enter the BIOS setup interface and set the system configuration 10 correctly. When the system completes the self-check of all hardware devices, it will list the hardware configuration list and resource allocation and occupation list of the machine on the screen. As long as the system can complete the self-check, it has been proved that most computer accessories have no fatal faults and their performance is basically normal. If the system still can't start, the problem should be the software. 1 1. Other abnormal conditions cause the system to fail to start normally.

1), the switching power supply is unstable due to internal filter capacitor leakage or capacitor loss, and its DC output voltage is high or low. For example, when+12V is low, it is easy to lose the hard disk, the system can't find the hard disk or the optical drive is not easy to unpack. If the machine can be started normally, the large-area hard disk is prone to bad tracks during use, resulting in data reading and writing errors or loss. If the voltage of +5V is abnormal, it is easy to cause failures such as the CD-ROM drive can't pass the self-test and the system can't find the hard disk.

2) Poor contact between the CPU and the cooling fan, or foreign objects between the CPU and the cooling fan, or the cooling fan is not connected to the power supply or there is too much dust. After starting, it is easy to appear that although the host can power on the self-test, there is always an abnormal alarm sound prompt; Sometimes you can enter the system desktop for normal operation; Sometimes it crashes at startup.

The rotation speed detection pin in the CPU cooling fan is broken, which causes the system to be unable to detect the rotation speed of the fan and start the machine.

The fault shows that the system suddenly loses power after a period of power supply. In fact, this is a protective measure to force the system to power off in order to avoid CPU damage when the CPU fan speed is found to be zero during the system self-inspection. However, it is often because of the sudden power failure during the start-up of the self-test, and sometimes the screen of the display has not appeared, so it is difficult to find the fault and find the fault location.