Exclude users of burning caused by overclocking the CPU, we have encountered in normal use of computer CPU processor failure rare. In general, if the computer does not start or is very unstable, we will from the motherboard, memory and other parts prone to failure to start the investigation, if the motherboard, memory, graphics, hardware and other accessories is no problem, then it is surely a problem with CPU .
Under normal circumstances, CPU very easy to judge after a failure, often have the following:
1, the system does not have any power to reflect, that is what we call the host point does not shine;
2, the computer crashes frequently, even in the CMOS or under DOS will crash situation. (This problem in other parts, such as memory, etc. This also happens after, you can use process of elimination to find fault source);
3, the computer to constantly reboot, the boot soon, especially the phenomenon of successive restart;
4, computer performance has decreased by a large extent.
Many of my friends to the CPU through the process of elimination to find fault, do not know how to rule out that the CPU after the failure, under normal circumstances would have replaced with new products. In actual fact, in many cases, as long as the CPU processor does not burn, we can solve various problems. Next to it in the form of real introduce several failure to appear causes and solutions.
One. Frequent crashes computer failure analysis and solution
Fault phenomenon: an Intel Celeron 1GB host, recently appeared in the course of the case of frequent crashes.Specific performance: boot into Win98 system can smoothly, but the use of 15 minutes or so, the system will crash,
Failure Analysis: dead memory, graphics card or motherboard problems and other accessories which can cause crashes, so the use of an alternative method, the various parts of the host one by one to replace him after the focus falls on the CPU. Through testing, found that the CPU core voltage of 1.2V, while the Celeron 1GB default voltage of 1.5V, the problem certainly lies in the processor.
Troubleshooting: As the default CPU voltage of 1.475V, 1.2V operating voltage is only now, so the cause of death is often caused by the computer is certainly due to lack of CPU power, in which case, probably because of the motherboard components, aging, resulting in low voltage power supply part, CPU naturally does not work, crashes also inevitable. Is the same as overclocking, upgrading the CPU frequencies are not stable, and some need to increase the voltage to stabilize at higher frequencies, this same reason, in fact, a considerable part of the computer and power failures are related.
Through careful examination, found on the motherboard CPU power circuit on a chip capacitors little yellow signs, then test it with a multimeter and found damaged, replacement troubleshooting.
Lack of CPU power for many reasons, first of all we should check that the power supply matches the machine, I encounter a lot of machine does not start or cause of instability is directly caused by the power supply. Check the power supply before we should first consider whether the computer is added a board and other accessories, some brand of power with the power system's total power is just the same, after adding the new board is likely to be caused by power of insufficient power supply. Therefore, if there is to add a new device may wish to try the device removed after, if not, then we must consider the stability of electricity supply in some parts of the unstable electricity supply will be such a problem, if it is you may wish to install a voltage regulator. Also too many USB external devices such as the installation will consume a large amount of power, can also cause system instability.
Similar problems have occurred, had a friend once bought a piece of card, returned home and found no light, it is estimated that many people now would think the problem is the graphics card, so they took back for, but at the dealer where the test a bit , actually no problem. Natural hapless friends who feel somehow, after tossing back and forth a few times, eventually chose a 300W power supply to solve the problem. Computer hardware problems too much, so we usually solve their own problems, it's best to think of several ways, thoughtful is the primary basis for troubleshooting.
In addition, because the CPU cooling efficiency, the situation may cause crashes. In addition to this failure may lead to frequent crashes under Windows, even in DOS or the board to enter into the CMOS may crash phenomenon. If you would crash under the CMOS, then it is probably due to the CPU causes the freezes. This case, the CPU can focus on examination of heat dissipation. When a failure occurs, they can open the case, detect the CPU temperature is very high, like the presence of hot phenomena.
Of course, this failure occurs, there may be high due to CPU overclocking causes the CPU voltage when pressure can not control, so that when the voltage range of more than 10% of the time, the CPU will have to increase the phenomenon of electron transfer, resulting in CPU internal injuries and crash failures, there will be severe burn CPU phenomenon. Therefore, we should not be too high for CPU overclocking, the CPU can be added with a powerful fan, might as well use other cooling measures to allow the CPU to better heat dissipation and thus solve the problem.
Second, the board continually reboot Troubleshooting
Fault performance: For the love machine safely through the summer, again in June I bought a CPU cooler, stable operation of the machine after installation for a month or so, because I have been using the computer's frequency is not high, and therefore did not encounter What is the problem. But with the increased use of the frequency and the weather heats up, the problem arises. After booting the machine can only work 40 minutes, and then is restarted, with the use of increasingly long time to restart the increasing frequency. Then lock the root causes of failure replacement radiator.
Failure Analysis: CPU heat generated can not be timely dispersed, but occur frequently occur due to temperature is too high and crash phenomenon. Under normal circumstances, if the host after working for some time the phenomenon of frequent crashes, we must first check the CPU heat dissipation.
Troubleshooting: Now determine the source of the problem with the radiator in the boot of the case see the functioning of the radiator fan, everything is normal, the fan is no problem. So the radiator re-removed, through the understanding of the cleaning and re-installed, turned on the issue before. Then replace the cooling fan, all OK. Is the heat sink problem, finally found after repeated contrast, the direction of the original clip is loaded backward. Resulting in heat sink and the CPU core in contact with the gap, CPU overheating, motherboard detection CPU overheating, restart protection. The original CPU cooling fan installed properly, will cause Windows to automatically reboot or not boot.
CPU process and integration with the continuous improvement of core heating is a more serious problem, and therefore the current requirements of the CPU cooling fan is also increasing. Cooling fan problems caused by improper installation is quite common and frequent. If you are using a Pentium 4 or Athlon like the CPU, please select the excellent quality of the CPU fan, and must pay attention to the correct installation method. Noir is a light machine or reboot, re-Noir CPU burned.
In addition, if detected in the BIOS CPU temperature rises too fast, it may be a problem with CPU coolers, also, or not properly installed. High temperature electronic transport phenomena occur, thus shortening the CPU life. Is under 53 ℃ for CPU temperature is too high, and prolonged use will cause system instability and hardware damage.
Three, CPU frequency automatically drop down failure analysis and scheduling
Fault phenomenon: the normal use of the computer, after the original 1.6GM-Hz power of the CPU into 1GMHz, and showing some "Defaults CMOS Setup Loaded" the message, re-enter the CMOS Setup to set CPU parameters, the system would normally display 166 frequency, but over a period of time then there have been - on failure.
Failure analysis ruled out: this failure is common in soft-set CPU parameters of the motherboard, which is due on the motherboard battery supply, so that the CMOS settings parameters can not be saved due to long-term effective, - like the time when the failure occurs , replace the battery on the motherboard can be solved.
In addition, the temperature is too high can also cause a sharp decline in CPU performance. If the computer is exceptionally stable performance during the initial period, but then a significant decline in performance, sometimes accompanied by the phenomenon of death. So if you use anti-virus software without killing found, use the Windows Disk Defragmenter is useless sort, format reinstall the system is still not, then please open the case to replace the radiator.
Equipped with a thermal control system of the processor, it will continue to sense the temperature. As long as the core temperature reaches a certain level, the system will reduce the processor's operating frequency, until the core temperature to a safe limit below. This is the real cause of system performance degradation. At the same time, it also shows the importance of the radiator, radiator recommended priority of some brands, but they also have hierarchy, the purchase should pay attention to its highest frequency of the CPU can support what is, and then fill a prescription according to its own CPU each other.
Equipped with a thermal control system of the processor, it will continue to sense the temperature. As long as the core temperature reaches a certain level, the system will reduce the processor's operating frequency, until the core temperature to a safe limit below. This is the real cause of system performance degradation. At the same time, it also shows the importance of the radiator, radiator recommended priority of some brands, but they also have hierarchy, the purchase should pay attention to its highest frequency of the CPU can support what is, and then fill a prescription according to its own CPU each other.
Fourth, caused by poor contact pin failure analysis excluded the host can not start
Fault performance: in 2004, assembled a compatible friend, use the Intel Celeron 1.7GHz processor, with all the normal two years, the latest move in a will, the love machine can not start up. Fan is running, CPU temperature is not high, also tried replacing memory, can not solve the problem.
Troubleshooting: After pressing the power supply chassis, the machine does not have any reflection to bring a certain amount of troubleshooting difficulty, no recourse but to use substitution to find where the problem lies. So borrowed from a friend's computer about a configuration, starting from the power, with the substitution by step troubleshooting. Replacement of a friend of the CPU, the machine starts normally, run the test program three hours did not exhibit unstable behavior. The authors is the machine's processor is broken, should not, this machine has no more than frequency, and friends do not often use, and the only office applications is online. In addition, CPU fan speed has been normal, which also shows the processor should be no problem. So my CPU feels that she inserted in a friend's machine.After all the normal boot, run very stable.
This is strange, CPU and motherboard, no problems, but this board can not recognize my original CPU, which in the end is how the case? So carefully observed two CPU, a 1.7GHz, a 2.0GHz, the same as the 478-pin Celeron, but it seems that there was no different. After careful observation, suddenly found the pins on the CPU has a slight clip marks. Then think, is not due to the CPU socket and CPU pin caused by poor contact between the fault. But CPU socket 478 feet, bad feet which it? Consideration to a friend's piece of CPU can run properly on my machine, this bad should be relatively minor and, therefore, my CPU into the CPU socket, and press down gently on the CPU socket wrench, when the feeling of pulling With some resistance when inserted CPU, the CPU forcibly from the outlet (there is a certain danger in this action, attention socket wrench can not be held down tightly, pulling up the CPU when the vertical force, or bent CPU pins is very troublesome a). Repeat several times, and then reload the CPU, fan wear, boot, everything is normal.
Fault Summary: This failure is not common, because the CPU pins are gold plated, but according to Intel's requirements, CPU sockets also need gold, because gold conductive properties, resistant to oxidation, so between the CPU and socket contacts are not prone to bad situation. But with the intensification of price competition in the market, there are a lot of the CPU socket on the motherboard and not plated or gold-plated thickness less than the required thickness of Intel, so in use, over time, CPU socket is easy to produce a layer of oxide layer between the CPU and socket appear bad, cause the machine to crash and so on. The method I use is the CPU socket pin with the friction, thereby undermining the oxide layer on the socket, so you can make contact between the CPU and socket changed for the better as far as possible.
Five, CPU overclocking failure analysis and solve difficult
Fault performance: last year to buy AMD Athlon 3000 + processor, Kingston 512M × 2 DDR400 memory, Gigabyte NF4 chipset, Seagate Barracuda 160GB SATA hard drive nine generations, Colorful 7600GT graphics card, the problem did not work properly, but can not overclocking the processor, which makes the author feel very depressed.
Failure analysis and solution: read this configuration, the processor still has a very high overclocking capability, there is no problem over to 2GHz. Configuration of the other accessories are brand-name products, should not be overclocking the main screen of the neck, will not be a power problem, right? As the parts used in the chassis comes with power. Although the nominal power 300W, but in the test, 30 minutes after the test, the test results from the figure, found that the 5V power supply has a serious side load, voltage fluctuations up to 5.24V, almost reached the 5V side can bear limit (5V reasonable fluctuation range: 4.75V ~ 5.25? V), while there have been some 12V voltage increased up to a 12.2V.It seems really not very good this power, load capacity is poor, OCCT Moreover, the parameters I have set too conservative, if transferred to Highest maximum CPU usage, but even difficult to stick to the test. ?
So then carefully observed this no-name power supply that this serious "shrink" the phenomenon, its output power of each port is too small: 12V ~ 6A, 5V ~ 13A, 3.3V ~ 5A,-5V ~ 0.5A ,-12V ~ 0.5A, the output power to 130W or so, and God knows this figure there is no water, and if so, then it simply can not support high-power AMD processors, let alone overclocking! Then replace the power supply, CPU FSB immediately stable over 166MHz, but you have to add on a 200MHz 0.05V voltage. But this has made me satisfied, it seems really overclocking failed power at play, such a good CPU was almost "wasted".
Seize the time to market to buy a brand is rated at 300W of power, put to test again, this time directly transferred to the FSB 250MHz, restart the machine after a smooth start, and run a variety of games, applications are no problem, now, the problem of the murderers get to the bottom.
No comments:
Post a Comment