How to fix a PC at home. Computer repair tips. The most important reasons for poor computer performance
Hi all. The last couple of weeks have been a lot of work, so the blog posts suddenly stopped appearing.
Now that I have free time, it’s time to do a pleasant and useful thing - blogging. Today we will talk about repairing your computer yourself. Which can be done not only for free, but also better than some masters. Modern world It’s hard to imagine without a personal computer or PC. We are used to the fact that computers surround us everywhere - at home and at work, in stores and other various institutions.
Wherever we are, everyone is working on computers. We use them for work, write diplomas and term papers, read interesting books and find out the latest news, play various computer games and watch interesting programs. And most importantly, we have the opportunity to communicate with our favorite friends scattered around the world.
However, it happens that such an irreplaceable thing as Personal Computer, fails. Do not rush to contact the service center specialists. In addition, the price of future repairs may slightly discourage you.
Let's try to understand the computer, find out and diagnose its breakdown, and try to repair the computer ourselves. Well, if it doesn’t work out, then we will always have time to contact a service workshop.
In addition, each user is able to carry out simple repairs on their PC themselves, without resorting to the help of specialists from a service center.
Get rid of jumps electrical voltage online will help you network adapter or UPS - uninterruptible power supply.
It won’t be difficult to purchase one; read more about how to choose a UPS in one of the following articles so you don’t miss it. Power surges electric current networks can most often arise when electrical appliances are incorrectly connected to the network, when you receive a large unevenly distributed load on one power line - for example: you plugged into one outlet through an adapter: a microwave oven, a heater and a kettle, which individually each take a lot of voltage, and together - a lot.
And if you also connect a PC, you will probably get that very sharp surge in electrical voltage that will damage your computer. It is very rare, but quite possible, that the transformer box may have a general breakdown, or a very strong wind will “sway” the wires and antenna, and so on.
Static stress in the human body may seem insignificant and almost invisible to the eye, but it can also affect the performance of your computer.
To remove it, simply place your hands against the battery or something iron, shake vigorously a couple of times, and after that you can safely begin repairing your PC with “clean”, non-electrified hands.
If the power supply is working properly, it is quite possible that the power button does not work when you turn on the computer. To check its functionality, you need to remove the power button connector and close it with a screwdriver. If the computer's power does not turn on, then the power button is not the cause of the malfunction of your PC, but most likely it is the power supply.
By shorting the power button, you may have the cooling fan turn on for a second. There may be several of them in a PC, and only thanks to them does case dust accumulate inside the PC.
Since complex electronics do not like dust, do not forget to clean and periodically blow out your PC. A vacuum cleaner, a can of compressed air or a soft watercolor brush are suitable for this. Read more about how to properly clean your computer in the article -.
1.1 The computer turns on and then turns off
In this case, the fault may lie either in the PC's motherboard or in its processor.
Capacitors located on the motherboard serve to rectify the overall voltage. When they overheat, they fail, losing capacity. To prevent this from happening, it is important to ensure that the PC does not overheat.
Don't throw away the motherboard right away. Due to overheating, it gradually fails. The computer freezes and " blue screen death,” and only after a few months the computer’s motherboard may completely fail. Then you will change it. More about critical Windows errors read the article - BSoD errors.
The fact that the PC often turns off may be due to insufficient cooling of the processor. And it is likely that the computer's processor is fine.
In this case, having removed the static voltage from the hands in a way known to us, we disconnect the PC from the power supply, remove the cover, and carefully get to the processor located under the large radiator and large fan-cooler on the motherboard of your computer.
Disconnect the three intertwined wires of the cooler fan. Carefully remove the cooler from the radiator. Try not to use a screwdriver when doing this. Most likely, the cooler-fan is attached to the radiator with a clamp and can be easily removed.
Carefully removing the radiator with cooler, we will see a metal box, the lid of which we also carefully open by pressing the metal tab. And very carefully remove the computer processor.
If the processor is burned out, you will see it immediately. Take a careful look at where the processor is installed; perhaps one of the legs is bent.
One option is to try to straighten it with a thin needle, but only as a last resort and last resort. Purchase thermal paste from a computer store that transfers heat from the PC processor to its heatsink. Using thermal paste, you can eliminate all the tiny gaps between the heatsink and the processor.
First clean the computer processor of old thermal paste and only then apply new one. Carefully return the processor to its old place, then install the radiator in its place and connect the cooler.
Try turning on your PC. If the computer does not turn on, the processor needs to be replaced. When replacing an old processor with a new one, be sure to lubricate the new computer processor with thermal paste.
1.2 The computer turns on and beeps
The PC turned on, but beeped. And this indicates a malfunction of some part of it. We all know that when turned on, the computer makes one short beep. This shows us that all the components and parts of our PC are working properly and the initial test has been successfully passed.
The set of PC fault melodies is different. Their purpose depends on the computer manufacturer and its BIOS model. It's easy to find out. To do this, find the BIOS chip on the motherboard.
If the chip is marked AMI, see the table given in this. If Award, look at the following table given in the same article. If the inscription on the chip is different, you will need to check it on the Internet.
To do this, enter the name of your BIOS chip on the motherboard of your computer in a search engine.
1.3 Overheating of computer components
One of the most common reasons why a computer constantly reboots is overheating of the computer's processor or video card. Always monitor the working condition of the fan-cooler located on the PC video adapter.
You can view the values of temperature sensors using the program that I talked about in the article -.
During the summer heat, it is better to refrain from playing energy-intensive games on your computer. Initially, the appearance of a blue screen of death or artifacts on the screen is associated with the inability to fix an error in the kernel of the operating system itself.
If any executable program or computer driver wants to execute an impossible command. You can try to eliminate the blue screen by simply restarting your computer. But if you do nothing, the error will return again.
The sound of a computer being turned on if one of the cooling coolers is humming is reminiscent of a running tractor. The computer cooling system with fan-coolers wears out, the seal in the bearing becomes thinner, dust and dirt get there, which leads to inevitable vibration and humming.
In this case, find out which fan is making this sound and replace it with a new one. This turns out to be quite simple. The cooler of the computer power supply practically does not hum, which means that the cooler remains on the video card, processor and case.
You have identified a humming cooler-fan and decided to eliminate it, which means that you no longer feel sorry for such a worn-out cooler and you can make an attempt to revive it. Then carefully remove it from the radiator with a small Phillips screwdriver and place it in front of you on the table.
1.4 Remove the cooler and try to clean it
The place where the wires exit is usually covered with adhesive tape. After removing this film, see if it is possible to disassemble the cooler fan into its component parts. If this is possible, then we carefully do it. The cooler plug can be removed using a small flathead screwdriver.
We release the axis with the fan blades. The dirty white piece of plastic that appears is the worn-out remnant of our fan gasket. This is what causes the loud and harsh sound of the computer.
Moisten a cotton swab with cleaning agent or cologne, and clean the fan axis and the place where it will be inserted in the future. Now it’s time to make a new gasket.
To do this, use a stationery knife to cut a small piece from the core of the gel pen. The resulting thin ring will be your new gasket for the cooler fan axis.
Use a small drop of machine oil or similar as a lubricant. Now put the axle in place with a new gasket, secure everything with a plug.
At correct installation old fan-cooler into your computer, you will no longer hear an unpleasant sound when plugged into the network.
1.5 Slow startup and computer performance
The computer is turned on, but the system boots too slowly. The folders required for work “load” for more than a minute, as does the launch of other programs on the computer.
Most likely the hard drive is to blame - such a malfunction in the PC is typical only for it. Perhaps your hard drive is dusty, just try to carefully blow it out. Keep it sealed - no need to disassemble.
If this does not help, then you will need to fully optimize your computer. I described in detail how to do this in the articles:
To be on the safe side hard drive It's always better to have a computer backup copy all your data on a removable external hard drive connected to a computer via a USB connector or on a separate disk in the form of an image created using a special Acronis programs. The capabilities of which I described in the article - .
It is best to store photos and other personal files and the operating system reinstallation program on an external optical drive, in the original box in a convenient place in plain sight.
General computer operating tips
1. When finishing work on your computer, try to turn it off from the power supply. This will protect your PC from sudden power surges, which happen quite often. The fact that the computer's power button is turned off will not save it from sudden changes in voltage in the electrical network.
Because of power surges, my computer's power supply, printer and active speaker have already failed once. Now I don’t step on the same rake.
2. It is better to leave the computer running all day than to constantly turn it on and off. This adversely affects his working resource.
This is primarily due to increased wear and tear on computer components, the elements of which expand at high temperatures and contract at low temperatures (when the computer is turned off). Therefore, computers in the gaming room, where they work day and night, last much longer than in your home, for example.
3. The computer does not tolerate sharp impacts on the case. At best, this will shake up the internal system dust.
At worst, a heavy heatsink for the central processor may fall off from a rough blow, which will lead to a temperature jump inside the case and failure of the PC processor itself, as well as damage to the video card and its connector.
The hard drive also cannot withstand sudden shaking and can easily break. And in total, computer repair will cost you a lot. By the way, the hard drive must be firmly fixed in the computer case. This will protect it from vibrations during operation.
4. Try not to drink any drinks while on the computer. Accidentally spilling liquid on a computer case can damage your computer.
5. When working with a computer, use an anti-virus program. Virus programs - Trojans, as a rule, zombify your computer to hack some of your files.
This means your PC will be hacked without your permission and knowledge. It won't actually harm your computer. If your PC gets a virus that can disrupt the computer’s control algorithm, then most likely something in the computer’s operation will fail.
If you often have problems with viruses on your computer, I advise you to read the section -. It contains articles with which you can easily remove the most evil versions of viruses.
7. When going to the store to buy a new computer part, be sure to take with you a broken old part to compare the part you need to replace your computer. This will make it easier for you to determine what exactly you need. It is quite possible that you are offered to buy something that is completely unsuitable for your computer model.
2 Conclusion - DIY computer repair
This article outlines the main points for repairing your computer with your own hands. If other faults are detected, it is better to contact computer repair specialists or service center, where all the equipment necessary for computer repair is available.
Each PC element and part has its own resource reserve. It is not at all necessary that after the allotted period for operation of a PC part has expired, it will fail. This durability is typical for older computers.
As a rule, with proper operation, only the fans need to be replaced. However, their performance is several times less than that of some modern mobile phones.
In the future, we will consider more complex repairs of computer components, such as repairing the power supply, for example.
- so as not to miss. That's all, may your computer serve you much longer than you expect :)
Most hardware problems are related to poor connections or minor glitches. In this case, the hard drive will be fine, that is, your files will not be affected. Back up the contents of your hard drive. Do this before you begin repairing your computer. To do this, remove the hard drive from your computer, connect it to another computer via an IDE-USB adapter (or a SATA-USB adapter for older hard drives
) and create a backup copy of your hard drive on another computer. Check the power cable.
- Double check that the power cord is connected to the computer and to the electrical outlet, and that the outlet is working. Plug your computer directly into an electrical outlet to see if this is causing the problem. network filter
- or extension cord.
If you are using a laptop, make sure the power adapter is securely connected to it. Try a different monitor.
If your computer is on but the screen is black, there may be something wrong with your monitor. Double check your monitor connection and then try connecting a different monitor (if possible). Remove the battery from the laptop and connect the power adapter to it.
You can work on a laptop without a battery if you connect it to a power source. If the laptop starts without a battery, there is something wrong with the battery that will need to be replaced. Check the power supply cables.
Make sure the cables coming from the power supply are securely connected to the motherboard. . Check the power supply
Old power supplies often fail, but they are fairly easy to test. A faulty power supply is the most common reason why a computer won't start. Look for loose screws.
A loose screw can cause a short circuit on the motherboard. Gently shake the computer case and listen for metallic rattling sounds. Use your fingers or long tweezers to remove the screw from the computer case. Inspect the cables.
Look for cables without insulation, as exposed wiring can cause a short circuit. Replace such cables. Reinstall the components. Try removing and then reinstalling the video card and modules random access memory
- You can also try to reinstall the processor, but this is a difficult task and most likely the processor is not the cause of the problem. Moreover, when reinstalling the processor, it can be damaged - in this case, the computer will not work at all.
Remove the discrete graphics card (if equipped). Now connect the monitor to the built-in video card (on the motherboard). A faulty video card can cause your computer to not start.
- If necessary
Repairing your computer yourself? Is quite real. Here we will work specifically with computer hardware, using measuring and installation tools. Or we’ll try to replace individual computer components ourselves.
DEAR READERS! If you are not confident in your abilities, you should not start. Due to certain and understandable circumstances, not everyone is given the opportunity to independently understand the malfunctions of your electronic friend. Consider again if it's worth it...
If you are going to delve deeper into the physical repair of a computer, laptop, monitor, etc., you cannot do without theoretical preparation. Repair specific devices, examples of which will be discussed here, will most likely turn out to be useless and incomprehensible for your specific case. So look around the site, I will make links to articles on theoretical training. I'll try to chew it thoroughly.
Further. The main focus will be on recovering information from hard and floppy drives. This is the most significant place in the matter of repair. Storage media are not expensive, but the data stored on them is of great importance to the user.
If we are talking about a stationary option, then I am inclined to the problems of repairing only its individual blocks. Blocks that make sense and can be repaired at home. Specific blocks:
- power supply and troubleshooting problems due to impending or detected failure
- replacement of computer components. Repairing microcircuits (from the motherboard to RAM sticks and replacing bridge chips) makes sense if the computer components are really expensive. And in this case, you have a direct road to the workshop. Save time, money and nerves. But with the help of my lessons, you will not allow yourself to be deceived by the price of the issue and the repair time. If the cost of the board does not exceed 2000 thousand, then believe me, sometimes it is better to save a month on beer and buy a new copy.
- monitors. Separate column. CRT monitors are gone. Modern screens with their electronic filling no longer allow us, home-grown repairmen, to dig around with our old-fashioned soldering irons with thick tips and solder that is quite dusty and gray with age. Real fighters now benefit from soldering stations of considerable cost, which allow them to work with small radio components. So I will limit myself to demonstrating the repair of large electronic attachments.
- printers and other office equipment. I practically don’t do it, but for the sake of replenishing my knowledge base, let’s consider this issue.
List necessary equipment I don’t bring it in for repairs on purpose. Each example has its own and will be given separately.
Repair of equipment. Theory. Tool. Measuring instruments.
With your own hands
Windows installation
Repair of equipment. Practice.
Details Created: December 20, 2009 Updated: May 11, 2015 Views: 19376DIY computer repair
Probably every PC user encounters the problem of a computer not turning on or glitching sooner or later. However, do not rush to take the system unit to the warranty department or workshop, or buy new components: in most cases, diagnostics and repair system unit carried out without much difficulty in a couple of hours with your own hands.
Typical mistakes of novice users.
The story about “malfunctions” and “repairs” should perhaps be preceded by a short lyrical digression on the topic “man and computer”, dedicated to what novice users usually encounter. Having heard on the phone that:
- "The computer won't turn on"- keep in mind: in most cases this means that the computer is simply deprived of the electricity necessary for its operation. This may seem funny, but advice on how to correct a situation with a “non-working” computer should be given to a beginner sequentially, starting with the simplest questions: “is there electricity in the sockets (are other computers working)”, “are the computer and monitor plugged in”, “ Is the surge protector and unit turned on? uninterruptible power supply" and, finally, "is the monitor turned on, and is the power switch on the back of the computer in the right position." I have met quite a few people (including one computer science teacher) who regularly forgot about such "little things", and once myself I made a mistake at this point: I thought that the system unit had failed, but it turned out that the cleaning lady accidentally pulled out the monitor’s power cable.
- "When you turn on the computer, it writes something about a faulty disk and does not turn on"(the user forgot to remove the floppy disk from the computer and looks at the message “Non-syststrong disk or disk error...”). Less common are keyboards and mice accidentally disconnected from the system unit. Remember the old BIOS joke about “Keyboard not present, press F2 to continue...”?
- "The computer is acting strangely". Advise you to check if one of the Shift keys on the keyboard is stuck, which is sometimes almost invisible in appearance, but has a truly magical effect on the operation of the computer. Sometimes beginners press buttons on the keyboard unnoticed by themselves - for example, they reach out to some “especially distant” key and accidentally touch others (Alt and Ctrl and spacebar are the most likely candidates). There are even semi-anecdotal stories about beautiful ladies who managed to bend over while working on the keyboard so that they pressed the space bar with their own bust. Jokes aside, of course, but teaching people how to use a keyboard and organizing a comfortable workplace is really necessary.
Standard preventive measures.
Regardless of why the computer is not working, it doesn't hurt to follow the standard sequence of actions (which, generally speaking, is good to do regularly).
Electronics is the science of contacts. Back in the 50s of the last century, when even control circuits, primitive by today’s standards, looked like hefty heavy chests with electrical equipment, as Boris Evseevich Chertok recalls in his wonderful memoirs “Rockets and People,” failures of control systems were caused by connectors that failed at the most crucial moments gave designers a lot of headaches; and if you believe that modern high-frequency and much more complex and richly connected computer electronic circuits have gone far ahead in this regard, then you are mistaken. This is still one of the most vulnerable places. Therefore, if something suddenly happens to your computer, first of all make sure that all peripheral devices are actually connected to the connectors and are securely seated in them. Sometimes undocking and re-docking helps. A "faulty" USB device often starts working in another USB connector. Just do not forget that PS/2-, COM- and LPT-devices can burn the corresponding port when the computer is turned on. Therefore, before moving anything in the external connectors (not to mention the internal ones), be sure to turn off the computer.
Problems often arise with the highest frequency and “multi-pin” slots for memory and graphics adapter. The memory strips must be securely secured with latches at both ends (since they fit into the slots with great effort, sometimes one latch does not close completely, and the memory eventually comes out of the slot). If the memory modules are secured correctly, and there is a malfunction, try moving them left and right and forcefully pressing them against the PCB board (a temporary measure, however, I perform this operation on my own computer every six months). It's also a good idea to wipe the contacts on the memory module with alcohol (just let the alcohol dry before putting the modules back into the computer).
Video cards sometimes have heavy coolers, the weight of which can literally cause the card to fall out of the graphics slot under certain conditions. Therefore, carefully secure it on both sides - both on the case and with the latch on the graphics slot in which it is installed. Otherwise, the recommendations are the same as for RAM: rock the board from side to side, press on it so that it fits deeper into the slot, wipe the contacts with alcohol. IDE and especially SATA connectors are fixed on the motherboard and on devices it is far from as reliable as we would like. By pulling out such a connector and then putting it back in place, it is sometimes possible to resolve a problem with a hard drive or optical drive. Just do not mix up the connectors, otherwise, for example, Linux in this case may simply not boot.
Various thin wires leading to the buttons and indicators of the case, although they do not fall out on their own, are very easily pulled out. And some wires (for example, to the speaker) assemblers are completely lazy to connect. Therefore, it won’t hurt to check for “orphan” dangling wires either.
Almost any system unit after six months of operation turns into some kind of dust collector. Dust deposits lead to a sharp drop in the efficiency of fans (up to their complete stop) and air cooling systems (dust is an excellent heat insulator), which gives rise to a whole set of glitches. Turbine-type fans, installed on many powerful video cards and in some models of CPU coolers, become particularly clogged; it is imperative to regularly clean them from dust!
Don’t forget that the power supply also has a fan (and sometimes more than one), which also becomes clogged with dust. However, unlike the system unit, disassembling the power supply is not only quite difficult, but, generally speaking, it is prohibited (as is clearly evidenced by the sticker, the rupture of which leads to the termination of the warranty). So it is better to “vacuum clean” the outside of a moderately dirty power supply unit. Such “prevention” takes five to ten minutes and often helps eliminate the problem.
Power is supplied to the system unit, all connectors are checked, the system unit is cleaned of dust, but the computer is still acting up? This means that the problem is serious enough that you will have to start looking for it. Let's start by dividing all faults into three large classes.
"The computer won't turn on."
This class of faults boils down to the fact that the computer cannot load any operating system, including safe mode MS Windows and MS-DOS boot floppy. A simplified description of the problem from the “user side”: a black monitor screen, no visible reaction to user actions, a beeping sound from the computer or a scary error message almost immediately after turning it on.
Since the procedure for loading the operating system is completely standardized, we will consider it, analyzing possible problems at each step.
Every computer starts with a power supply. Despite the widespread belief that a power supply is a box with radio elements that passively converts 220 volts to +3.3; +5; +12 and others like them, this is not so. Since the days of the ATX standard, the power supply has been a fairly intelligent independent subsystem of the PC.
Under “normal” conditions, when voltage is applied to the computer input (plugged into a power outlet and the power supply switch is turned on, if there is one), the power supply begins to produce a special “standby” voltage, sufficient for low-power computer chips. It is required, in particular, so that you can turn on the computer from the keyboard, by local network etc. So it should be taken into account that a number of elements remain energized (albeit weakly) in the case of a working power supply, if the computer is connected to an outlet. Therefore, the standard recommendation for any manipulations with the internals of the computer is to turn off the system unit from the network (or turn off the power supply with a switch located on the rear wall of the case).
On some modern motherboards, near the power connector there is a tiny LED indicating the presence of this very standby voltage. If your motherboard has it and doesn’t light up when the power supply is turned on, it means the power supply is faulty (optionally: the fuse is blown). Just don’t confuse it with the Power Good indication LED, which is also quite common; it is usually larger and lights up when you turn on the computer (check your motherboard manual for details). If there is no LED, then the presence of standby power can be checked with a tester.
When you press the power button on the case, the motherboard transmits a signal to the power supply via the PS On power cable. Having received it, the unit begins to actually “turn on”, going into operating mode. Since a fairly large load is connected to the power supply (hard drive spindles, fans are spinning up, other “mechanics” are starting), the voltages produced by the power supply at this moment can fall quite short of the declared ratings or simply “jump” - in general, there is a typical transient process, taking approximately 0.1-0.5 s. A special control circuit monitors this process and, after reaching the mode when the power is within the “acceptable limits,” it issues a signal via the Power Good line, which serves as a “green light” for the motherboard to start operating the “smart” electronics.
If the power supply does not show signs of life after pressing the power button (does not work or the fan stops immediately), it means that either the button is not connected or does not work (try manually closing the PWR contacts in the group of front panel connectors on the motherboard with a screwdriver), or the power supply is not capable of providing sufficient power required by the system (try disabling optical drives, hard disks and video card and repeat the process), or the power supply is faulty. If a computer with Award BIOS encounters problems at this stage, the speaker may beep continuously.
If everything goes well, then after pressing the power button, all the computer fans start (including the case fans and the power supply fan), and the hard drives spin up the spindles [However, for SCSI drives and some IBM/Hitachi models, you can set a “delayed” start - for example, indicating that the first disk in the array should start immediately, the second - after a second, the third - after two, etc. Time staggering is necessary because SCSI disks, as a rule, have very powerful “mechanics” (10- 15 thousand disk revolutions per minute), consuming a large current “at the start”, and there may be a dozen or two of the drives themselves, and even a powerful power supply is unable to “pull” the simultaneous start of such a combination. But in ordinary user desktops, “delayed launch” is exotic].
Any power supply must have a short circuit protection circuit built into it, which automatically turns off the unit in the event of a short circuit. If the power supply refuses to start, then perhaps it is not the fault, but the burnt-out “mother” or one of the peripheral devices; in the block, the built-in short circuit protection simply triggers. We check whether this is true by turning off the peripherals one by one and trying again to start the computer.
The power supply quality is monitored continuously. If, for example, the voltage in the electrical network drops for a second, which leads to a drop in the voltage supplied by the power supply, this circuit, according to the standard, is required to remove the Power Good signal from the motherboard, which, according to the standard, is equivalent to pressing the Reset button on the computer. If there is a #PG signal (+5 V on the gray wire) and the computer does not start, most likely the power supply has nothing to do with it (or is of extremely poor quality [In cheap power supplies, manufacturers sometimes skimp on the power control circuit, but the Power Good line simply shorted to the +5 V line). If the computer reboots under heavy load, and the voltage supplied by the power supply is lower by more than 5-10% of the nominal value, then, most likely, the power supply simply does not have enough power to “pull” the system. The power of power supplies decreases over time [The power remains the same, the electrolytic capacitors simply dry out and the filtering of pulses along the power buses deteriorates. As the power supply ages, the power limit at which these loads begin to affect the operation of computer components decreases. - S.L.], and sometimes the unit not only cannot withstand the next upgrade, but also begins to “drop” the previously stable system. If you have any suspicions, check the power supply using the appropriate S&M utility test (see below) or try disconnecting unnecessary optical drives and hard drives to see if the problem goes away after that or not.
Power supplies, especially inexpensive ones, are by no means omnipotent when it comes to converting the ugliness that lives in our electrical networks into stable +5...+12 V. Surge filters are practically useless here (at best they will protect the computer from such “nice” events such as a falling off neutral wire in the riser [Which gives stable ~380 V for several hours in the entrance electrical network. Half of the neighbors burn out half of their imported electronics after such an emergency, and a burnt-out surge protector is still nicer than a burnt-out monitor or computer]) - use. uninterruptible power supplies!
Many power supplies (especially old ones) are not able to function normally without a load. If you are planning to turn them on without the motherboard assembly, connect at least a hard drive [An “empty” power supply can be turned on by connecting the green wire to any of the black ones on the main ATX connector. - S.L.].
Anyway, The best way To make sure that the power supply is faulty is to replace it with one that is obviously more powerful and of higher quality. If everything works, then the power supply is to blame. Just remember that a good three-hundred-watt unit often carries a much larger load than a low-quality four-hundred-watt unit.
If everything is fine with the power supply, then after receiving the Power Good signal, the motherboard removes the Reset signal from the central processor, as a result of which the processor conducts a self-test and begins to carry out the computer initialization procedure stitched in the BIOS. If the processor is faulty, then visually nothing happens - the fans are spinning, the hard drives are spinning, voltage is supplied to the PS/2 keyboard and mouse, but there are no other visible signs of life. To make sure that everything is in order with the processor, unplug the processor cooler fan connector and make sure that after a short time the cooler heatsink gets very hot. If the power supply is working properly and the processor does not heat up, we draw the appropriate conclusions. In addition, you should check whether the north and south bridges of the chipset, MOSFETs in the processor power circuit, as well as the chips on the RAM modules are heating up. They should all be at least warm to the touch (just be careful with the “feeling”!); and if nothing heats up, then either the motherboard or the power supply is faulty; If a separate element on the board does not heat up, then either it or the motherboard is dead.
If the CPU is alive, then the first thing the BIOS launched from ROM memory does is carry out the procedure for checking and initializing system components (Power On Self Test, POST). The chipset, DMA and memory controller are initialized one by one. If the RAM exists and works, then the BIOS copies itself there. The keyboard controller is then initialized. Then the BIOS scans the system address space in search of the video adapter BIOS, checks its checksum and, if there are no errors, transfers control to it. That, in turn, tests and initializes the video card (at this moment a corresponding splash screen appears on the monitor screen), after which it returns control to the main BIOS. If the video card BIOS is not found, the system BIOS tries to initialize the video card on its own using standard methods.
Up until this point (video card initialization), any boot problems are reported to the user using sound signals speaker. If nothing appears on the monitor screen and there are no sound signals, check whether the speaker is connected (on many motherboards it is soldered on the board itself, on others you need to independently connect the case speaker to the corresponding contacts on the board). Particularly advanced models of motherboards can also use an integrated sound module as a speaker (even reporting errors in a human voice); sometimes an LED or segment indicator is soldered or connected to indicate the completion of POST (for details, see the motherboard manual). If the speaker is connected, but there are still no signals, it means the processor or RAM is faulty, or the BIOS is damaged.
After the video adapter is initialized, all peripheral devices connected to the system buses are sequentially searched for and initialized (corresponding messages may appear on the screen). Loading custom BIOS settings. The advanced capabilities of the central processor are initialized (in particular, the cache memory of the second and third levels is turned on), and information about its manufacturer, model and frequency is displayed on the screen. Most BIOSes produce one short beep after this initialization, indicating that everything is fine with the system. At this point you can try to log in for the first time. BIOS Setup by pressing the Del (or F2) button.
Then the BIOS alternately checks the RTC (Real Time Clock) timer, tests the RAM, initializes devices on the ISA bus (if one exists in the computer), initializes IDE controllers and the FDC (Floppy Disk Controller). The screen displays a list of PCI and ISA devices, connected optical drives and hard drives, their characteristics and other information about the hardware configuration. This completes the preparation of the computer for operation, and the BIOS looks through the devices specified by the user in the settings one by one and tries to boot from them. If a floppy drive is indicated in the list of boot devices, it comes to checking it and there is a floppy disk in it, which, however, does not fit the standard boot disk, the BIOS displays the traditional message “Non-syststrong disk or disk failure...”. Other devices are ignored if there is no boot record. When a suitable boot entry is found, the BIOS loads it from the media and transfers control to it. The boot record usually contains a system bootloader, which, having received control, loads itself entirely into RAM, asks the user which operating system to boot, and launches the bootloader of the corresponding OS. From this moment on, the computer can be considered to have started.
If any troubles arise when going through the POST procedure with an already initialized video adapter, the BIOS reports them in plain text. Some BIOS hide the POST screen behind a proprietary graphical splash screen, but it can always be disabled in BIOS Setup or by pressing a specific button. Remember that although POST usually “flies” in seconds, by pressing the Pause (Break) button you can try to pause this procedure in order to familiarize yourself with the messages displayed on the screen (which will help you at least understand what has been determined in the computer in general).
How to use all the above information? Turn on the computer. After a short time (about half a second), a splash screen logo of the video card should appear on the screen, or a sequence of sound signals describing the malfunction should sound. If there is neither one nor the other, make sure once again that the computer is plugged in, the power supply is working, the power button is connected to the motherboard, the processor is heating up, and the speaker is connected. If it doesn’t help, start undocking all peripheral devices one by one (remembering to turn off the computer, or better yet, unplug it from the network). Remove each expansion card one at a time, ending with the video card. Unplug all peripherals and cables one at a time, ending with the mouse and keyboard. If it doesn't help, remove the RAM sticks. If, after removing another device, the computer starts working (or beeps that something is missing), the culprit of the malfunction has been found. If the processor is heated alone on the motherboard with a connected speaker, but stubbornly refuses to squeak and work, then the motherboard itself has almost certainly failed.
Having determined in one way or another (by the squeak of the speaker, by test messages or by a special POST indicator) where the boot process “collapses”, we begin to figure out what happened to the corresponding device.
On motherboards, as a rule, the VRM (Voltage Regulation Module) fails - a circuit that provides the most important components of the computer with the required power (in particular, converting +5 or +12 V from the power supply into 0.9-1.6 V consumed by the central processor). Visually, it looks like a palisade of high-capacity capacitors (vertically standing cylinders, if anyone doesn’t know), interspersed with chokes (inductors) and power switches - MOSFETs (black or gray square, relatively large hanging elements on the motherboard, sometimes covered by a radiator) . The circuit is located next to the processor socket, chipset, expansion slots and memory slots. Since modern electronics are very power-hungry and at the same time low-voltage, a very large current passes through the VRM, especially the processor one (up to hundreds of amperes) - on modern motherboards for Intel processors The area allocated for the VRM and processor power cables literally emanates warmth. Provide normal functioning so powerful circuit- a whole problem, and it was especially pronounced on motherboards of the Athlon XP and the first Pentium 4 generation, when the current consumed by the processor grew at a faster pace than the improvement of VRM. Most often, electrolytic capacitors burn, and MOSFETs often burn out (a burnt-out power transistor is noticeable by the charred PCB around it). Less often, the clock generator (a small rectangular chip in the area of the north bridge of the chipset and the processor socket) fails. Sometimes a metal object such as a bolt gets onto the motherboard, causing a local short circuit, or the motherboard shorts out to the case. Dismantle and gently shake the board to make sure that nothing is stuck there, place it on some antistatic bag like the one in which it was supplied, and run it outside the case, connecting the power supply and shorting the contacts of the connector for connection Power buttons screwdriver
Today, according to my statistics, one of the first places is occupied by failures caused by defects in electrolytic capacitors of power filters (in power supplies and on motherboards). The computer often remains operational, but regularly glitches. This is especially noticeable in load changing modes: accessing disks, optical drives, USB devices, heavy load on the processor, etc. Visually, this defect can be identified by the swollen upper part of the capacitor. Moreover, the failure of one capacitor usually leads to a chain reaction - all its colleagues connected in parallel also die (due to the fact that the smaller the total capacitance, the greater the heating of individual capacitors). Alas, high-frequency interference on the power bus, where the capacitances have died, is not caught by anything, not even by an oscilloscope, since it is very short-lived. (Often, such a failure is mistakenly identified as a lack of power in the power supply - if you install a power supply with a larger output filter capacity, the glitches partially disappear. But in fact, you just need to change all the swollen capacitors, this is not at all difficult if you know how to hold a soldering iron. - S.L. )
Failure Motherboard BIOS board failure most often occurs due to failed flashing. It would seem, what is the probability of a loss of electricity in those two minutes that the firmware lasts? One millionth? And strangely enough, I have seen the consequences of such a “coincidence” once - Murphy’s law has not yet been canceled. Often the BIOS is mistakenly loaded with firmware that does not work with the given motherboard[Some companies have particularly “succeeded” in this with their sophisticated labeling system. It says one thing on the board, another in the file... And all BIOSes are incompatible. The firmware program does not determine that the board is not the same. - S.L.] - use, if possible, online flashers that download from the site themselves the required version, automatically detecting the type and revision of the board, or at least check more carefully what you are going to flash!
However, the range of problems with the BIOS is not limited to flashing - for example, many nVidia chipsets were famous for successfully crashing the BIOS if they crashed (due to overclocking, for example) or lost power during POST.
What to do if the BIOS really screwed up? Some motherboards have a disaster recovery function, implemented either as a second BIOS chip or as a BootBlock routine - an area of BIOS ROM that is erased or rewritten by a separate procedure (different from writing the BIOS itself). Perhaps, to restore the system, you need to press certain keys on the keyboard or rearrange the jumpers on the motherboard - look at what the user manual that came with the motherboard says about this. If the manual has long been lost or there is nothing about the procedure BIOS recovery is not described - try connecting a floppick to the motherboard and carefully see what happens to it when you turn on the computer. If the motherboard is dead, but the floppick is being accessed (the light comes on even for a second), then to bring the BIOS back to life, just prepare a special “recovery” floppy disk, insert it into the drive, turn on/restart the computer and wait about five minutes (computer screen it remains black, but the floppy drive can rustle like a floppy disk). For details and a “recovery” floppy, please visit the motherboard manufacturer’s website.
By the way, do not forget to check that the BIOS reset jumper (Clear CMOS) is in the Normal position. Many motherboards will not start if the jumper is in the Clear position.
If you are unlucky and cannot restore the BIOS using built-in means, then, unfortunately, it is impossible to do without outside help. If the BIOS chip is simply soldered onto the board (which, I must say, is not uncommon), then the only way to fix the board is to unsolder the chip, reflash it, and solder it back. If the microcircuit is installed in a special “crib”, you should first make sure that the contacts are reliable (it’s unlikely, but it’s still possible that it’s all about them), and then either remove the microcircuit from the “crib” and take it to the nearest owner of a suitable programmer, or use the brutal method of “flashing on the fly” - take a known working BIOS chip from a friend from the same motherboard, tie it with a thin strong thread (so that you can easily and quickly pull it out of the “crib”), install it on your motherboard instead of the non-working one, turn it on computer, wait full load, prepare everything for flashing, then, without turning off the computer, pull out the neighbor’s BIOS chip, put your own in its place and start the firmware process. The trick is based on the fact that access to the chip mainly occurs only when the computer boots, exactly until the moment it becomes possible to copy its contents into RAM, after which the computer “forgets” about it. However, I will not recommend such a risky method, unless it is very far from the radio market and repair shop.
A broken processor is usually not fatal. Under the influence of high temperature, voltage and current, electrical transitions in the crystal gradually degrade and fail, and the newer the processor and the finer the technological process by which it is produced, the higher the temperature and voltage, the faster this happens. Keep in mind that the dependence of the reaction speed on temperature is exponential (an increase in temperature by 10 degrees leads to an increase in speed several times) and depends significantly nonlinearly on voltage, so if you heat the processor to a couple of hundred degrees Celsius or exceed one and a half to two times the nominal voltage - the chip will fail in seconds and fractions of a second, and if not observed temperature regime and raise the voltage relative to the indicators stated in the specifications, then instead of the two to five years of service promised by the manufacturer modern CPU, you can easily bring him to the grave in a few months.
On processors with an open chip (most Pentium III, all Athlon and Athlon XP) mechanical damage crystal - chips (even tiny ones, sometimes almost invisible) - be extremely careful when installing the cooler and do not drop the system unit on the floor! Naturally, neither electrical nor mechanical damage to the crystal can be repaired.
If the RAM does not work, try cleaning the contacts with alcohol and make sure (I repeat again!) that there is normal contact between the modules and the slots. If you can get into BIOS Setup, try reducing the memory frequency and increasing the timings. A good option for inexperienced users is to use the Fail-safe defaults item or something similar. If you cannot enter BIOS Setup, try resetting the BIOS settings (using a jumper or the Clear CMOS button on the motherboard) - the BIOS will load the most secure settings from its point of view, and you should still be able to enter BIOS Setup. If this technique does not help, remove all memory modules except one. If a module works alone, check the other modules one by one. If it doesn't work, try inserting it into other slots (don't forget to reset CMOS!). If it does not work in any of the slots, then it will never work in this system. If it works, try adding other modules one by one. Unfortunately, it may be impossible to make all modules work at the same time - then leave the working combination, go to BIOS Setup and install auto mode determining memory frequencies and timings. If this also does not help, forcefully set the memory characteristics even below those values that the BIOS considers safe. Remember that dual-channel memory operation is not possible in all combinations of modules in slots, so you should start searching through these combinations, switching to single-channel only if dual-channel options do not work.
Video cards are the same motherboards in miniature, with their own processor and VRM circuit. Make sure there are no swollen or charred hanging elements; if the video card has been dropped (or generally subjected to mechanical stress) - check for chips on the crystal GPU. But first of all, check the contact between the video card and the motherboard again, as described in “preventive measures”! In general, it’s better not to have any special hopes that a really broken video card can be brought back to life on your own. A burnt or chipped GPU will have to be resoldered (which, firstly, is not very trivial given its standard BGA packaging, and secondly, a working GPU is required, and they are not sold in the store), and, for example, a microcrack in the board that has arisen because the cooler is too heavy, it is generally impossible to fix it, except to find the poor fellow with the same faulty board and see if it is possible to assemble one “alive” from two “dead” ones.
In general, it is difficult to say anything about repairing expansion cards. As usual, check and clean the contacts; try moving the card to another slot; If it doesn't work, take it in for repair.
Well, don’t forget to make sure that the diagnosis given to the device is correct! “The video card does not work” - put another one in its place, put the “non-working” video card in another computer. Maybe it's not the video card, but the motherboard that's at fault; or there is incompatibility of two specific components, etc. Just keep in mind that inoperative components can damage other components. A faulty processor can destroy a VRM in a split second, a faulty power supply can easily burn half of the components; I even heard about a motherboard with a faulty IDE controller, which once a month destroyed another hard drive. Handle parts known to be faulty with the greatest care!
"Obvious computer malfunction"
Unfortunately, successfully passing POST does not guarantee that all devices identified in the system will function correctly. Let us first examine the case when the corresponding malfunction is obvious, that is, it is completely obvious that the device categorically refuses to work normally.
Most often, optical drives fail. The service life of their mechanics today rarely exceeds two or three years, so if the drive stops reading CDs, spins them many times, starts writing with errors, or opens the tray on the twentieth attempt, it’s time to run for a new one. Unlike hard drives, optical drives can be completely disassembled with your own hands, so if you wish, you can try to shake out all the dust that has accumulated over a couple of years, clean, lubricate or replace worn-out plastic mechanics. Before installing a new drive into the system, be sure to check that the Master/Slave jumpers are set correctly. Sometimes they forget to connect the power cable to the drive or do not insert the power connector and interface cable tightly enough.
Hard drives fail much less frequently, but, unfortunately, if they do break, they often take with them all the data stored on them into oblivion. Therefore, if problems arise with the HDD, you will have to act with extreme caution. First, remove the disk, write down its name and try asking Google about it. If you are lucky enough to have a disk like IBM DTLA or Fujitsu MPG, then you will learn a lot about the technology of its manufacture and the mistakes of the developers, as well as what is happening with the hard disk now, what will happen in the near future and how it should be repaired so that it can work for at least a couple more days. If detailed description there is no problem on the Internet, you will have to act on your own, which in itself deserves a separate article. But in short, then...
Gently shake the drive - no noise should be heard from it. extraneous sounds. Judging by the sound, some small parts dangling inside the HDA almost certainly shouldn’t be dangling there [There are exceptions. They concern laptop hard drives, especially Hitachi, where there are “normally loose” parts in the parked position of the heads. - S.L.], so if the data stored on this disk is still valuable to you, it is better to immediately take the disk to a workshop.
Unplug the interface cable and turn on the computer. The characteristic smooth sound of a spinning spindle and the chirping of the heads should appear for a few seconds (this is undergoing initialization and initial calibration), which are then replaced by the smooth hum of a running engine. If, instead of this soft melody, the drive grinds, whistles, howls, vibrates, it means that its mechanics are catastrophically damaged (the engine is faulty, the disks are misaligned, the heads have gone beyond the limiter due to an impact). There was nothing to drop on the floor! The absence of any sounds most likely indicates an electronic failure. Check for cracked and charred elements and microcircuits on the controller board (connecting a hard drive to a running computer does not lead to the same effects), make sure that the HDA is securely connected to this board. If you know how to use a tester, check using the HDA contacts that the spindle motor is working properly (has a resistance of about 2-3 Ohms). If you can hear that the engine is trying to turn on, but the pancakes do not spin up, the mechanics have failed (for example, the disk heads were not properly parked and “stuck” to the working surface - try to slightly “twist” the switched off drive in the plane of the disks to “unstick” "them, and if this helps, urgently save the data: our “patient” will not last long [Such disks work for me for a very long time. You just run it through MHDD, the disk reassigns several sectors to backup ones and lives on no worse than any other. - C .L.]). If you hear that the engine is running, the “pancakes” unwind, but instead of the rustling of the heads, several clicks and taps are heard, after which the engine stops (or does not stop, but the heads continue to tap evenly), then the head positioning system has failed. Immediately turn off the drive and do not even try to repair it yourself, but bring it in for repair! There they will be able to remove at least part of the data from it and, perhaps, bring it into at least a working state. Just keep in mind that new hard the disk will cost you significantly less than a drive “reanimated” in this way.
If the drive is mechanically sound, but BIOS Setup does not detect it (don’t forget to use the auto-detection option!) or produces some nonsense instead of the normal data indicated, by the way, on the HDA cover, then the controller chip is dying (on hard drives Fujitsu MPG, for example, due to insufficiently tested “eco-friendly” sealant used in the production of the controller, this sooner or later happens to all disks) or service routines recorded on the same “pancakes” as user data are damaged. Make sure that the interface cable is connected securely, switch the HDD to “Master” mode, and disconnect all other drives and optical drives. If this does not change anything, take the disk to a workshop or find an exactly the same working drive, temporarily move the controller board from it to your disk and, if possible, copy the data. This operation requires certain skills: it is not always possible to find a drive identical to yours, and small differences in the HDD revision when rearranging boards can lead to death. It’s better to find out in advance about the possibility of such rearrangements for a specific pair of disks (for example, contact specialists on well-known forums).
If the drive is detected in the system, but is practically inoperable, look for the MHDD utility from Dmitry Postrigan on hddguru.com/ru and start thoughtfully reading the documentation for it. Since MHDD works directly with IDE/SATA controllers, it makes sense to try to install and run it, even if the drive refuses to recognize the BIOS (just don’t connect mechanically faulty drives this way!). The utility runs in bare DOS; The images of boot floppy disks (or CDs) it needs can be found in the same place as the program itself. You will be very tempted to use a utility like Norton Disk Doctor, but don't give in to it - MHDD will not only tell you an order of magnitude more information about the state of the disk, but will also recover much more data. If the disk SMART along with the read test shows that the HDD is slowly dying, it is perhaps better to buy new disk, although compared to a mechanical or controller failure, you are generally lucky, and the drive will probably last a couple more years even with such threatening symptoms. However, storing important data on it is too risky.
Like any technology, floppy drives also break down over time. I won’t give advice on how to fix them - if you really need a floppy, you’ll somehow find ten bucks for a new one.
External ports, as mentioned at the beginning of the article, can burn out if they are handled incorrectly (connecting PS/2, COM and LPT devices to a working computer; short circuits; lightning strikes near the “air”, etc.) . Mechanical failures of frequently used connectors are also possible. Connect to another port, and if there are no ports left, buy an expansion card with a corresponding controller (they usually cost around $10) or an external docking station (up to $100).
There is nothing else seriously broken in the computer, although due to strong overclocking it can become so glitchy that it will not even be able to load Linux or Windows. For example, “the inability to load hal.dll or ntdll.dll” for NT-like Windows is usually associated not with a faulty hard drive, but with slightly glitchy RAM.
Unstable computer operation, also known as glitches.
I'll bet ten to one that any single computer glitch is associated not with hardware, but with software. So if a 3D modeling program, video codec or toy is glitchy, make sure that other renders, codecs and toys are also glitching before you hang all the dogs on the system unit; It is very likely that this is not the reason. If the problem manifests itself more or less independently of the programs used, it’s time to take up diagnostic tools [Oh... I just repaired my “mother”, whose drying out electrolytic capacitors affected a single program - the TWAIN driver of the Canon Lide 500f scanner. And nothing else! - S.L.].
The first and main program that will help us in this difficult task is the S&M utility, the latest version (1.7.6) of which can be freely downloaded from the developer’s website at the link www.testmstrong.nm.ru/snm.htm. Being well known in narrow circles under the name testmstrong, S&M not only inherited the wonderful properties of its DOS predecessor for testing RAM, but also acquired a number of new ones.
So, we install the utility, turn the testing time slider to “long” or even “infinite” mode, check the boxes next to the hardware that we are going to test, go through the test settings with the user manual in hand, setting the most stringent conditions, and run the test of the hour like two. Most often, the cause of minor glitches is RAM, and the best remedy there is no way to diagnose its failures.
In addition to S&M, download from mbm.livewiredev.com and install another free utility, Motherboard Monitor 5 (MBM5 for short), which integrates perfectly with S&M and allows you to monitor the voltages produced by the power supply and the temperature of the main system components recorded by the thermal sensors of the board, processor and hard drives. Unfortunately, the year before last the author quit working on the utility, but I still don’t know anything better. MBM5 requires rather non-trivial setup, so don’t be too lazy to import settings for common motherboards suggested by the program when you first start it and select your motherboard (or at least one very similar to it) from the resulting list. If your board is not in the list or MBM5, after selecting a similar model, does not start showing plausible numbers from the system sensors, try selecting the necessary sensor settings manually. I personally almost never monitored the negative voltages produced by the power supply properly, but they are still practically not used in the system unit, so it is not necessary to configure them, but here are the readings of sensors that monitor the temperature of the processor and other components, as well as the voltage on the lines +3.3 V, +5 V and +12 V - worth debugging.
With MBM5 installed and configured, we restart S&M, turn on the processor performance test and begin to monitor the increase in its temperature. If it tends to levels of about 90-100 degrees, after which the system collapses, you have serious problems with CPU cooling. If the temperature does not exceed 80 degrees, we wait for the end of testing of all processor subsystems (cache memory, FPU) with the checkbox checked to check the correctness of the results. If no problems arise, run the PSU (Power Supply Unit, also known as power supply) test in S&M and watch how the voltages on the +3.3 V, +5 V and +12 V lines creep down. If 12 V does not turn into 11 , and the system unit does not undergo an unexpected reboot - the software crash test can be considered successfully completed.
However, do not rush to jump with happiness to the ceiling, even if long-term testing in S&M did not reveal any faults. Download the Prime95 utility (www.mersen-ne.org/freesoft.htm) and run a stress test from it (torture test) in several suggested options. Prime95 itself is a client of the distributed computing project for searching Mersenne numbers, but its ability to catch the smallest hardware failures of the processor-chipset-memory combination is simply phenomenal. A complete check of system reliability using Prime95 takes almost several days, but, as a rule, the utility reports about hardware faults no later than half an hour later. If all tests pass without failure, you are looking for a hardware problem where you should be looking for a software problem; if Prime95 crashes, especially if it does not crash immediately, but after an hour or two of operation, the system is unstable, but it is quite possible that in practice you will not notice this instability, although the computer will slow down slightly. If P95 crashed on the first test, you have tested something wrong with S&M, since the system is openly glitchy.
The hard drive is checked for functionality, and at the same time it is treated using the already mentioned MHDD program. Unfortunately, it only works under bare MS-DOS, so at worst you can use as an alternative one or another traditional “disk doctor” like the one from the Norton Utilites kit.
To fully test the health of the video card, you will need to install some utility for monitoring and configuring the video card. Personally, I like RivaTuner the most latest version(www.guru3d.com/rivatuner), - with its help you can perfectly track the dynamics of the temperature of the graphics processor and video card (the temperature threshold here is higher than that of the central processor, but it still shouldn’t go off scale over 100 degrees), and also try to reduce clock speeds GPU and video memory, as well as disable various advanced features used by the video card. If, after switching the accelerator to low-speed and safe mode, the problem disappears, it means that the culprit has been found. In other cases, try removing the existing video card drivers and installing the new ones recommended on the forums (more recent or, conversely, more old version). If the “meeting did not help,” try temporarily replacing the video card with a similar one that is known to work, or putting the suspicious video card in another computer and checking it there.
What to do if suspicions of a hardware malfunction are confirmed? In case of problems associated with overheating (and this is most of them), disassemble the system unit, vacuum it and, without closing it, turn it on to make sure that all case and system fans are spinning. Some particularly smart pieces of hardware regulate the rotation speed of the fans until they stop completely, but in any case, when you start the computer, they should work for the first few seconds. Radiators of processor coolers, north and south bridges (even if the bridges are not covered by radiators), video cards should not burn when touched. Microcircuits on hard drive controller boards and microcircuits on memory modules should also not be too hot. MOSFETs in the processor power circuit on the motherboard are allowed to heat up much more, but still within reasonable limits. Under no circumstances should there be a burnt smell coming from the system unit.
If any element overheats, take care of its better cooling. Opposite hard drives that are too hot, place additional fan; cover the overheating chipset bridge with a more powerful radiator or even a small cooler (not forgetting about hot-melt adhesive or thermal paste); Remove clumps of dust from fans and radiators and drop some engine oil into the bearings. Sometimes it is easier and cheaper to replace a stopped fan than to repair it. If a fan that was dangling unattended in the casing jammed a fan, then the latter most likely did not suffer from this, but the connector to which the fan was connected could easily fail. If this happens, use adapters that allow you to connect the fan directly to the power supply: you will have to forget about monitoring the cooler speed, but at least the impeller will spin again. In the case of a processor cooler and video card cooling system, it can be useful to replace the thermal interface with a more efficient one (a high-quality AlSil-3 is ideal). Sometimes the processor cooler is simply not fully secured (or the fastening has become loose, as happens on Glacial Tech coolers). Just be more careful with the removal and installation of cooling systems: do not chip the fragile CPU and GPU crystals by negligence!
If the problem is not overheating, try “tweaking” the equipment settings towards lower frequencies and greater stability. Depending on what exactly is “buggy”, reduce the frequency of the CPU, GPU, regular and video memory, increase the timings, increase the voltage on the corresponding elements (just don’t overdo it: the voltage should not be increased by more than 10-20%).
In general, setting up a computer correctly is a very difficult matter, so if you are not confident in your abilities, it is better to contact a professional. If my warning did not frighten you and the methodical tweaking of the parameters in the worst side did not lead to anything - the equipment will most likely have to be changed. Check your computer for the typical hardware faults discussed in the two previous sections, since swollen capacitors on the motherboard do not always lead to its inoperability, but they definitely affect stability for the worse.
Precautionary measures
Finally, let me remind you that when carrying out repairs and diagnostics, a number of precautions should be observed.
You should connect and disconnect any hardware inside the system unit, as well as some peripheral devices, only when the computer is de-energized. Turned off the computer, connected or disconnected hardware, turned on the computer again. It is strictly forbidden to do this on the fly without turning off the power, unless, of course, you want to be left without some of the connected devices and lose some of the functionality of the motherboard.
If a warranty is provided for the entire system unit, it may be sealed, and violation of the integrity of the seals will lead to loss of warranty. Even if there are no seals on the case itself, they may be on individual elements - say, on the processor heatsink. Opening the power supply and optical drives also leads to loss of warranty for these components. Under no circumstances should you open the hermetic block of the hard drive. However, the presence of seals does not mean that the only thing left to do is take the computer to the warranty department - it is often possible to diagnose the machine and fix the malfunction without violating the warranty.
For God's sake, do not climb inside a power supply connected to the network without a clear idea of what you can touch and what you absolutely cannot touch! And don’t forget that computers, like any electronics, are afraid of what we carry with our hands static electricity. Wearing a special antistatic bracelet on your wrist, recommended by foreign guidelines, is completely optional, but taking off clothes on which static accumulates, taking off your slippers (if we are talking about home repairs) and touching the central heating radiator before work will not hurt.
Remember: you perform all manipulations with the computer at your own peril and risk. If something burns out when flashing the BIOS using the “pull out” method, don’t say later that I didn’t warn you about the potential danger of such actions.
Good luck to you in the difficult task of troubleshooting your PC!
Standard color coding for power supply wires.
Black: ground, GND, COM
Orange: +3.3 V
Red: +5 V
Yellow: +12 V
Pink: +5 V Stand-by ("standby" power supply)
Blue: -12V
White: -5V
Green: Power supply is on
Gray: Power Good signal (PWR_OK, +5±1 V)
Typical Front-Panel Connectors Color Coding
Blue: Reset button connector
Light green: Power button connector
Red: one of the contacts is marked with a cross: HDD Led indicator connector
Green: one of the contacts is marked with a cross: Power Led indicator connector
Orange: cabinet speaker connector
If your desktop computer has become unstable or has stopped turning on, check the documents to see if its warranty period has yet expired. Before contacting a warranty workshop, you must make sure that everything is in order with the software and that the computer is not infected with viruses, since malfunction of the computer for the above reasons is not a warranty case and you will have to pay for this expensive service.
Some programs can restart the computer after an update if there is a checkmark in the settings that allows the computer to be restarted without notification after the update. If after the audit installed programs The computer continues to work unstably, which means that the reason lies in a hardware malfunction and repair of the material part will be required. This is a warranty case and it is better not to repair it yourself.
To prevent the computer from malfunctioning, any connections of connectors and blocks can only be made when the system unit is de-energized; you need to turn off the pilot or remove the plug from the 220 V socket.
The main reasons for computer instability
Malfunction of coolers (fan)
One of the most common computer breakdowns is the failure of cooling coolers due to a decrease in their speed or stopping. The processor heatsink can also become clogged with dust. So if everything is in order with the programs, then you need to first check the operation of the coolers.
Power supply failure
The second most common cause of computer malfunction is the failure of the power supply unit (PSU) of the system unit. A malfunction of the power supply may manifest itself as obvious or not obvious.
If there is an obvious malfunction of the unit, the computer does not show signs of operation when turned on, the coolers do not work and the LEDs do not light up. If there is a non-obvious malfunction of the unit, unstable operation of the computer, sudden system freezes, spontaneous reboot, errors during operation.
The power supply is the most loaded part of the computer and in addition is exposed to instability in the power supply network in the form of voltage surges and natural ones during a thunderstorm. Therefore, it is recommended that during a thunderstorm, it is mandatory not only to turn off the computer, but also to remove the electrical and network Internet plugs from sockets. You can work on a laptop during a thunderstorm provided it is powered by a battery and connected to the Internet via Wi-Fi.
Malfunction of random access memory (RAM)
It is not uncommon for a computer to become unstable due to a faulty RAM memory. Even positive test results special programs cannot guarantee 100% serviceability of the memory, and reliability can only be verified by replacing a memory stick with a known good one. If there is more than one stick installed in the slots of the computer, then removing one at a time, you can thus perform a check. If the computer starts to work stably, it means that the removed memory stick is to blame.
Hard disk drive (HDD) failure
If the computer has been working for many years, then the hard drive (hard drive) is often the culprit for unstable operation of the computer. But, as a rule, if the hard drive malfunctions, even before the systems start to boot, a message about a disk read error is displayed. Sometimes the hard drive begins to make sounds uncharacteristic of its operation.
If the system boots and there is a suspicion that the hard drive is faulty, then you need to test it for bad sectors on the disks and set a ban on writing to these sectors. After a while, check again, if new bad sectors appear, the hard drive will have to be replaced.
There are programs that allow you to physically make a copy of the hard drive installed in your computer to a new one, of any capacity. Then you won’t have to waste time re-installing programs and drivers.
Start button malfunction
I also had to deal with a completely unusual reason for the computer to randomly shut down due to a malfunction of the Start button. The fact is that if you hold this Start button on the system unit pressed for more than three seconds, then force termination system operation and computer shutdown. So, the contacts in this button periodically randomly closed and thus turned off the computer. For a long time I could not understand what the reason was until I thought of disconnecting the wires coming from this button from the motherboard.
Microcracks in motherboard tracks
And finally, the saddest case is a disruption in the stable operation of the computer due to microcracks formed over time in the tracks of the motherboard. They appear on some models of poorly designed motherboards.
Microcracks appear as a result of attaching the processor heatsink not to the socket into which the processor is inserted, but to printed circuit board. Since the radiator is pressed with quite a lot of force, the motherboard bends and microcracks form on the tracks, which leads to periodic contact failure. First, the computer begins to freeze periodically and eventually stops working completely.
Perhaps this design of mounting the processor radiator is not an accident, but was done with intent so that after completion warranty period consumers bought a new motherboard or a new system unit.
I have repeatedly encountered such malfunctions of motherboards. In some cases, modifying the radiator mounting helped, after which the computer began to work stably. If, when opening the system unit, you discovered a similar radiator mounting system, then I advise you to redo it, using my experience.
If your computer freezes when loading a system program
BIOS (Basic input-output system) is the primary software that is stored in a special chip on the motherboard. When you turn on the computer, the processor first accesses the BIOS and reads the “guide” for further action from it. The task of the BIOS also includes checking all the main components of the computer.
When you turn on a working computer, if there is a speaker in the system unit, always at the beginning of boot software One short beep sounds. To this sound BIOS signal everyone is used to it, and no one pays attention to him. One short beep means that the automatic diagnostics of the performance of all computer units, carried out by the POST (Power On Self-Test) program embedded in the motherboard BIOS, was successful, all units are operational and ready for use.
However, if the computer's BIOS emits several short or long beeps, it will be difficult not to pay attention to this, despite the fact that in this case the software will not load. These signals are not random and by their combination and duration you can determine a computer malfunction.
How to determine the type of BIOS installed on your computer
In home computers, one of two types of BIOS is the most popular - Award or AMI, and for each type the same sound signal indicates different events. Therefore, to decipher the fault, you need to know what type of BIOS, Award or AMI is installed in the computer. At the first moment the computer boots, a hint is usually given at the bottom of the monitor screen which key must be pressed several times to enter the BIOS, it can be one of two “DEL” or “F2”, usually it is “DEL”.
When you enter the BIOS control panel, you will see something like this.
At the top of the screen the BIOS type is written, in this computer Award installed.
through the system speaker of a computer with Award BIOS
| Table of beeps emitted by the BIOS through the system speaker of a computer with Award BIOS | ||
|---|---|---|
| Sound produced | Decoding sounds | |
| 1 short | The computer is working properly | No action required |
| 2 short | Minor errors found. A message appears asking you to enter the BIOS | You need to check the BIOS configuration and resolve any inconsistencies |
| 3 long | Keyboard controller error | Check that the keyboard cord is securely connected to the system unit |
| 1 long + 1 short |
RAM (Random Access Memory) error | Check that the RAM sticks are installed correctly, remove them and put them back in, and move them to a free slot. |
| 1 long + 2 short |
Video card error | Check whether the video card is fully inserted into the slot on the motherboard and that the video card cooler is working |
| 1 long + 3 short |
||
| 1 long + many short ones |
Error reading from BIOS | Check the presence of a contact in the contact pad of the BIOS chip |
| Repeating series of short sounds | The power supply is faulty | |
| Repeating series of long sounds | RAM fault | Replace the RAM stick with a working one, or remove one of them if more than one is installed |
| A repeating series of short and long sounds | Processor failure | Check the functionality of the processor cooler and the reliability of the radiator mounting |
| The beep sounds continuously | Power supply failure | Check the output voltages of the power supply |
What do the sounds made by the BIOS indicate?
through the system speaker of a computer with AMI BIOS
| Table of beeps emitted by the BIOS through the system speaker of a computer with an AMI BIOS | ||
|---|---|---|
| Sound produced | Decoding sounds | Recommendations for troubleshooting |
| 1 short | The computer is working properly | No action required |
| 1 long + 1 short |
The power supply is faulty | Check the output voltages of the power supply |
| 2 short | RAM (Random Access Memory) parity error | Disable parity check in BIOS |
| 3 short | Error testing the first 64Kb of RAM | Increase the size of memory typings in the BIOS, if that doesn’t help, replace the RAM stick with a working one |
| 4 short | System timer malfunction | |
| 5 short | Processor failure | Check the functionality of the processor cooler, the reliability of the radiator fastening, remove and insert the processor, if it does not help, replace it with a working one |
| 6 short | Error initializing keyboard | Check the connection in the connector is secure. Click the button Num Lock, if the indicator on the keyboard does not light up, then the keyboard is faulty |
| 7 short | Motherboard failure | It needs to be replaced |
| 8 short | Video card memory error | Video card replacement required |
| 9 short | BIOS checksum is incorrect | Check the presence of a contact in the contact pad of the BIOS chip, flashing it may help |
| 10 short | BIOS write error | Check the presence of a contact in the contact pad of the BIOS chip, the error is not critical and you can continue working |
| 1 long + 3 short |
Video card error | Check that the video card is inserted all the way into the slot on the motherboard and that the video card cooler is working |
If the computer does not turn on
You click the Start button, but the computer does not turn on. In order to understand the reason why the computer is not working, you need to analyze its behavior after pressing the power button. Several options are possible.
Complete absence of any sounds emanating from the system unit and the glow of the LED on the monitor. It is unlikely that the monitor and the system unit failed at the same time. It is necessary to check the presence of a power supply in the electrical outlet, the serviceability of the pilot, if present, for which it is enough to plug the pilot into the outlet table lamp or any other electrical appliance. Are the plugs inserted into the sockets, are the power cords connected to the monitor and the system unit, are the switches on the monitor and on the back of the system unit turned on. Most likely, you will find the reason and the computer will work.
The battery is low
The computer may not start if the lithium battery, which is a flat cylinder with a diameter of 20 mm and a thickness of 3.2 mm and is installed on the motherboard, is discharged. As a rule, a CR2032 or similar battery of the same size is installed. It is worth noting that the battery marking contains its overall dimensions.
The battery's job is to save the settings made by the user in the BIOS memory when power supply is not supplied to the computer and keep the clock ticking. A harbinger of the end of battery life is the failure of the clock and date readings after the computer is turned on again.
Such a battery must be checked by measuring the voltage at its poles with a voltmeter. It must be at least 3 V. If the battery voltage is less than 3 V, then it should be replaced with a new one. If the voltage is more than 3 V, then you need to wipe the contacts in the battery compartment and the plane of the battery with alcohol. When installing the battery, the polarity must be observed. A new battery usually lasts more than five years.
To remove the battery, you need to remove the side cover of the system unit (how to remove the cover is described in the article below), find the installation location of the battery compartment and move the metal retainer aside. After removing the battery, you may need to restore the BIOS settings and set the current date and time.
Short circuit in the power supply circuit of the components
The computer is connected to the power supply, when you press the Start button, the power supply cooler impeller jerks, the impeller does not continue to rotate, and nothing else happens. This means that the power supply protection system is triggered due to the presence of an overload in the power circuit in any of the installed elements in the system unit or in the power source itself.
To determine the cause of the failure, it is necessary to open the system unit and sequentially disconnect the wires coming from the power supply, except for the wires connected to the motherboard. It's easier to start with disk drives. Before turning off the unit again, it is necessary to de-energize the system unit. Next, all cards are sequentially removed from the motherboard. After disconnecting the next device, they try to start the system unit. If the startup succeeds, it means that the culprit of the computer failure is the last disabled unit. If all units are turned off and the power supply does not start, it means that the source itself or the motherboard is faulty.
If you have the same power supply, even if it is of lower power, then you can temporarily connect it for testing without removing the standard power supply from the system unit, but simply disconnecting it from the components of the system unit.
From the power supply to the motherboard, supply voltages are supplied using a 20 or 24 pin connector and a 4 or 6 pin connector. For reliability, the connectors have latches.
In order to remove the connectors from the motherboard, you need to press the latch upward with your finger at the same time, applying quite a lot of force, rocking from side to side, and pull out the mating part.
To check the power supply, you need to completely disconnect it from all equipment, leaving only any that is powered by a four-pin connector or a Serial ATA connector, for example a hard drive or any disk drive.
If the connector has 20 contacts, then you need to connect the leads together 14 (the wire is green, some power supplies may have a gray wire, POWER ON) and 15 (black wire, GND).
If the connector has 24 contacts, then you need to draw up conclusions among yourself 16 (green, in some power supplies the wire may be gray, POWER ON) and 17 (black GND wire).
If the cooler works, then with a high probability the power supply can be considered serviceable. To be completely sure, the source should be checked on a load block or another working system unit, and the motherboard should be tested in a workshop.
If, when you press the Start button and are sure that the supply voltage is supplied to the system unit, the computer does not show signs of life, then the power supply is definitely faulty and you need to replace it or try to repair it yourself.
How to open the system unit
Before disassembling the system unit, it is imperative to remove the cable supplying the 220 V supply voltage from the outlet, since while it is inserted, the system unit is under a voltage dangerous to human life. When carrying out any work on the system unit, you must first remove the power cord and insert it last when connecting. In order to gain access to the contents of the system unit, you need to remove one of the side covers. Next, you need to unscrew the two screws (they are the top ones in the photo) that secure the cover. Push the cover towards you, the hooks will disengage in the grooves, and the cover will be released.
Visual troubleshooting of the system unit
After cleaning the system unit from dust and if the coolers are working normally, you need to carefully inspect the motherboard and the cards inserted into it. Particular attention should be paid when inspecting the appearance of electrolytic capacitors. In the vast majority of cases, the failure of the motherboard and power supply is associated with the failure of electrolytic capacitors.
Such a capacitor consists of two strips of aluminum foil twisted together through an insulating pad that is impregnated with electrolyte. This roll is inserted into an aluminum cylinder and sealed. From each strip a conclusion is drawn, one of which is connected to the plus and the other to the minus. Thanks to the electrolyte, the capacitor has a large capacity with small dimensions and is therefore called electrolytic.
In the event of an insulation breakdown, thermal energy is released, the electrolyte boils, and the pressure inside the capacitor increases sharply. To protect against explosion, notches are made on the end side of electrolytic capacitors. As the pressure increases, the housing swells or ruptures at the notch, and by this sign it is easy to find a failed capacitor. The main reason for the failure of capacitors is their overheating or exceeding the permissible voltage.
The photo clearly shows that the end of the left capacitor is flat, and the end of the right one is swollen and with traces of electrolyte leakage. This capacitor must be replaced. On the motherboard, electrolytic capacitors on the +5 V power bus fail more often than others, since they are installed with a small voltage reserve, only 6.3 V. I have encountered cases when all the capacitors on the motherboard on the +5 V power supply were swollen .
When replacing, I usually install capacitors along the +5 V power supply circuit, for a voltage of no less than 10 V, and along the +12 V supply circuit, at a voltage of 25 V. The higher the voltage the capacitor is designed for, the better, the main thing is that it fits in dimensions to the installation site. It is not difficult to replace a failed capacitor on the motherboard yourself.
If you have found and replaced swollen capacitors, you have most likely eliminated the cause of the unstable operation of the computer. You need to connect all the connectors and check its functionality. If appearance parts of the system unit available for review do not raise suspicions, then it is necessary to check the condition of the power supply capacitors. It's possible that you'll have to take care of it
Order 343 mail. Order by Russian post. Consequences of failure to appear in court when summoned
How to put a password on a folder on a Windows computer without and with programs
Pluton – Free Bootstrap HTML5 One Page Template
History of ZX Spectrum: Myths and reality New Spectrum
Voice assistant Siri from Apple Siri functions on iPhone 6s
How to roll back to a previous version of iOS?
Unlock iPad in four days