Troubleshooting for PC's

Following the care that we have seen over the previous threads, assemble computers is relatively simple. The vast majority of the components can be embedded only in one way and there are locks and grooves in the slots that prevent many combinations of incompatible components, such as a comb DDR fit on a board that only supports DDR2 modules.

Although the overall quality of the components is better today than it was in 2000 or 2002, for example, components with manufacturing defects are still a common occurrence. In English they are called "DOA" (dead on arrival), or are already sold defective due to damage incurred during transportation, or lack of quality control by the manufacturer.

Although they can be exchanged under warranty, in most cases without difficulty, they end up causing defective headache for anyone who rides micros, since beyond the time necessary to diagnose the problem, you lose time to change the number.

First of all, start with a checkout by checking the source of the connectors are firmly seated on the motherboard and the connector is installed P4 (most current cards will not boot with it unplugged), the video card and other peripherals are fully engaged, if all IDE and SATA cables are in order and if the buttons on the cabinet are connected to pins right panel of the motherboard.

Another tip is that many current cards will not boot if the power cord from the cooler is off, a precaution against damage caused by the processor overheating. If you are using a nobreak or stabilizer, try also try to connect the PC without it. It is not uncommon stabilizers very low quality or defective, providing an output so unstable that the source (for security) turn signal "power good" for the motherboard. Without the signal the board does not boot, as if the computer was unplugged.

After deleting the most obvious possibilities, the standard procedure to troubleshoot is to disconnect all non-essential devices, leaving only the plate processor (and cooler!:), One of the memory modules and (if the onboard video card ) also the video card. If the mike over to complete the POST, you can re-install the other components one by one, until you find the culprit. Often you will end up discovering that the problem was just barely touch on some of the fittings.

Also try to change the plates position. It is very common that a card is not detected when plugged into a slot, or even causes the computer simply stops the boot, but works perfectly in another. This can occur by defects in the contact slots, oxidation (in the case of computers used) or simply by poor contact to install the card in the first slot.

When doing maintenance on old PCs, it is also important to clean the contacts on the cards and memory modules, since oxidation of the contacts is a major cause of poor contact. Cleaning can be done using a rubber or vinyl, in the absence of one, using a ballot in good condition, has in his pocket. The paper money is abrasive, so it also works well as wiper contacts.

If, on the other hand, the computer will not turn on even the "minimum configuration", start by checking basics like the key source of 110/220 and the position of the Clear CMOS jumper. Also try cleaning the Setup settings, removing the battery and changing the Clear CMOS jumper position for 15 seconds. Another option to clean the Setup settings is to use a coin or screwdriver to close a short between the two poles of the battery, also for 15 seconds:

From there, the way is to go to trial, changing the memory module, source, video card, processor, and finally the motherboard itself, to find the culprit.

Although they are far from being an accurate indicator, the error codes of the BIOS, issued through the speaker may give clues of what is wrong in many situations. The error codes vary depending on the BIOS used, but these are the most common codes are valid for most of the cards:

1 short beep: This is the confirmation beep sounds when the POST is successfully completed and no error is detected.

no beep: If you turn the coolers, but the system does not boot and no beep is emitted (and you have checked whether the speaker is actually connected), we most likely a serious problem on the motherboard or processor. This also happens when trying to connect without the micro processor or no memory module installed. If nothing really works (the coolers do not turn, no sign of life) then the problem is probably the source (or stabilizer, or any other item of electrical equipment) or even the power button on the cabinet (you can have it the wrong pins on the motherboard, for example).

2 beeps: This is a kind of "general error", similar to a Windows blue screen. POST failed for an unknown cause.Here there is plenty to do besides using the method of trial and error to find out what is wrong.

1 long beep and 1 short beep: Problem on the motherboard (only generic error that indicates that something is wrong, without offering details.)

1 long beep and 2 short beeps or 1 bit long and three short: Problems with video card (or lack thereof). It is very common that the video card is ok, but poorly fitted or oxidized contacts. Try removing the card, clean the contacts using a rubber or vinyl one ballot and install it again. Most boards still boot after issuing the error, allowing you to access the computer via the network, but many do not really boot up you solve the problem.

3 long beeps: Error in keyboard. This error is relatively rare. He is not issued when the keyboard is not installed, but when it is present, but the driver is with a defect or short circuit.

2 (or more) long beeps: serious Problem in RAM modules. This error is generated only when the memory module is installed, but an error is detected during the first 64 KB, or when he passes the test of memory count of the BIOS. It is rare that a memory stick actually come to be damaged to cause this error (in module damage caused by static are usually more discrete, affecting only a small range of addresses within the module), the most common is that there is some unexpected problem of compatibility between it and the motherboard. First of all, test the computer using a different module (if possible another manufacturer) and test the module that caused the error in another board different. It is quite possible that both are good.

5, 6 or 7 short beeps: The processor is embedded, but with some serious damage. On some cards this error is also issued when the processor is overheating (the fan is connected loosely or is not installed, for example).

9 beeps: Error saving the BIOS, or damage to the flash memory chip where it is recorded. Nor is it a common mistake, because when you try to do a BIOS upgrade and it is unsuccessful, the card simply stops working, or start using some protection system included by the manufacturer. Physical errors on-chip Flash memory are quite rare.

Let a short list of symptoms related to defects in each component:

Video Card: 3D cards today are virtually complete computers, which have not only one processor (GPU), but also memory and power supply circuits themselves. Many cards also get energy directly from the source through a connector and molex or PCI Express 6-pin.

The most common problems with the video cards are defects caused by wear of the capacitors, as well as in the case of the motherboard. The defect starts manifesting in older games (when the power consumption of the board and consequently the stress on the capacitors is greater) and progresses gradually until the card actually stop working. In this case, it is possible that the PC simply does not work with the board stuck, the motherboard beeps transmitted pass, as if there was video card, or boot the computer normally, but no image is displayed on the monitor.

Video cards can also be free from defects in memory. If the defect is in the area reserved for the frame-buffer (more rare), you will notice the image corruption from the beginning of the boot. If the defects are in the main block, reserved for 3D rendering, you'll notice textures corrupted and in many cases also crashes during the games.

As with the processor, there may be cases of incompatibility between video cards and motherboards specific. This is relatively rare in PCI-Express we use today, but it was common when trying to install AGP recently manufactured in old motherboards, which had no power circuits rated to support them.

HD: Defects in HD does not prevent the motherboard perform POST. The PC starts to boot normally, but did not detect or can not boot your HD, for the start of loading the operating system. However, fit the IDE cable reversed (which is possible to use old cables, without chamfer) causes the computer to boot really leave. Defects in the flat cables cause corruption of data, which leads to symptoms similar to that of a HD with badblocks. When you encounter problems reading and writing to IDE drives, try first of all replace the cable.

Memory: There are many cases of conflicts between certain brands of memory and some models of motherboards.When the computer simply will not boot with a particular module (but works with others), or stops working after you install an additional module, it is likely that the module is good and the problem is simply an incompatibility between it and the motherboard.

Usually the manufacturers keep lists of modules tested, or manufacturers recommended, but they are not much use in practice, they are always incomplete. The best thing is not worrying about this and choose the memories based on technical characteristics, price and reputation of the manufacturer and replace the modules within the guarantee in cases where they do not work together with a specific motherboard.

Another type of defect, the most common addresses are damaged in the module. They can be caused by static, power surges, defects in source or power supply circuits on the motherboard, or even premature wear caused by the use of much higher voltages than the standard (for overclocking). This type of damage does not prevent the computer completes the POST and boot to boot normally. However, you will be faced with crashes and abnormal behavior of programs where the defective addresses are used by the system. It may be that defects are observed in the first address and the operating system crash during boot, or you may be at the end and problems are only noticed when running multiple programs simultaneously.

The best way to check the status of the modules is to use Memtest, which we saw in the memory. It is always important to let the test run for some time after mounting a new PC, because it not only detects errors in memory modules, but also the memory controller, trails and processor caches.

The Memtest can be found in many Linux distributions, as the Kurumin, Knoppix and Ubuntu (just using the "memtest" or "Memory Test" on the boot screen), is part of the utilities included in the Ultimate Boot CD (http:// www.ultimatebootcd.com) and is also available as an image. ISO in http://memtest86.com/.

Processor: It is very rare that new processors are defective. Quality control by manufacturers such as Intel and AMD are usually very good, so the chances of a processor out of the factory with defects is small. Damage during transport are also unusual, since the processor component is a small and very tough physically. Damage caused by handling, such as bent pins, are also easy to understand.

In any case, a defective processor or damaged by static can cause the computer simply will not start the boot (because the motherboard needs the processor to perform the POST), which is rare, or that the micro work, but present several stability issues.

Note that there are many cases of incompatibility between old motherboards and processors launched recently. There is no guarantee that a motherboard socket 775, manufactured in 2005 will be compatible with a Core 2 Duo, for example. It is common for boards need BIOS updates to support processors that were released after their manufacture and in many cases there are inconsistencies related to power supply circuits or chipsets on the board. In such cases, the system can not simply start the boot, as if the processor was burnt, when in reality it is the motherboard that is failing to initialize it.

Accumulation of dust: The whole process of cooling the PC is based on the movement of air (even when using a watter cooler you need at least two exhaust fans: one in the radiator used to cool the fluid and other power supply).This causes airborne dust to accumulate on exhaust, inside the source and other components. The accumulated dust affect the airflow and heat dissipation, which causes the computer may have overheating problems.

The classic symptoms are that the computer boots as normal, but after some time hang on or to perform heavy tasks.It is also possible that the dust close contacts between some tracks, causing several errors. The PC can go to make errors related to corruption of data stored in memory, for example, as if there was a physical error in the modules.

The solution is simply to make a good regular cleaning, disassembling the computer completely and using compressed air or a brush to remove any dirt. The power supply also accumulates too much dust, and can also be opened and cleaned. However, it is important to take care not to touch them, otherwise even with the power disconnected from the socket, the capacitors store a large amount of energy.