Resolving Memory Problems

Diagnosing the most common memory problems (Cont.)

When you attempt to install Windows 3.x, Windows 95, Windows 98 or Windows NT, the installation stalls, hangs or the system reboots itself.
Just a reminder, this is a memory troubleshooter, so we won’t go into all of the reasons why you may not be able to load these operating systems other than to say that we presume that you have (1) resolved all other hardware possibilities such as hard drive problems (and controller problems where necessary); (2) are aware of partition limitations for the specific operating system being loaded; (3) you know how to partition the hard drive, format it with the correct FAT and you have run Scandisk; and (4) you have created the necessary boot files to enable upper memory areas as well as load Smartdrv where necessary. This aside, here are some possibilities:

• Check the systems power supply
• Check for CPU overheat problems
• If you are loading Himem.sys in your Config.sys file, make sure that there are no errors occurring when this file loads.
• Swap the position of the memory modules and try the installation again.

Your computer boots, Windows starts but is unstable.
Instability in all Windows versions can be caused by any or all of the following:

• Power Supply problems
• A chip on a memory module is beginning to fail or is borderline.
• Parity of ECC error between memory modules
• CPU overheat problem
• Potential video card failure
• Irregularities with the voltage regulating module on the motherboard
• *If the motherboard is equipped with its own cache, it may be failing
• Corrupted Windows device driver or virtual device driver
• Corrupted Windows swap file (Page file in Windows NT)

You hear a continuous beeping sound from your system during power up or boot-up process.
A continuous beep tone from your systems motherboard is a sign of memory failure, failure of the memory on your video card or a complete video Card Failure. You may want to verify your settings in the BIOS to insure that there are no errors in the setup before changing the memory.

When you start your computer, it continuously runs a ram count without loading the operating system.
If, during startup (boot-up) your system continually runs a ram count, starts at zero and runs through the total amount of your RAM memory and then starts over again, this is an indication of any one of the following:

• Parity mismatch between memory modules
• ECC memory on a motherboard that doesn’t support ECC memory
• Mismatched memory modules
• Memory chip failure
• BIOS or CMOS setting error

No display other than a Windows blue screen error (GPF) on the monitor during boot-up.
Unfortunately this type of problem has many possibilities. The following presumes that you have tried to boot into Windows Safe Mode and cannot do so. You may want to check the following:

• Parity mismatch between memory modules
• ECC memory on a motherboard that doesn’t support ECC memory
• Mismatched memory modules
• Memory chip failure
• Incorrect CPU speed setting or CPU overheat condition
• A memory related BIOS or CMOS setting error
• Windows video driver problem
• A externally attached device such as a scanner is causing a problem
• Virus

No video display on the monitor of any type.
This presumes that the system is starting, e.g. the processor is running, but there is no indication of boot-up on the monitor. This is usually an indication of any one of the following:

• Memory failure
• Video card failure
• Incorrect BIOS settings or a corrupted BIOS
• One or more other system cards has failed on the PCI bus and is preventing the system from starting.

  You may want to remove all cards on the motherboard except for the video card and then attempt to start the system. If the system continues to refuse to boot, then swap memory module position and change video cards.

System hangs briefly at startup and/or reboots after prolong usage.

This can be a difficult problem to diagnose, especially if the problem is occurring randomly rather than on a consistent basis. Any one of the following, or a combination of them, will cause this problem:

• Power Supply problems
• A chip on a memory module is beginning to fail or is borderline.
• Parity of ECC error between memory modules
• CPU overheat problem
• Potential video card failure
• Irregularities with the voltage regulating module on the motherboard
• *If the motherboard is equipped with its own cache, it may be failing

You may want to run a memory diagnostic such as the one here to rule out a memory problem. You may also want to check your CPU fan to make sure it is clean and running as well as verifying that your power supply is adequate for the number of devices installed in your computer. You’ll find our calculator here.

All of the above are typical of, among other things, memory related failures. If you have tried all of the suggestions we have made, you made need the services of a well trained and knowledgeable technician to perform more advanced diagnostic methods.

Even when a memory error is presumed to exist, identifying the defective module is not an easy task. Given the nearly endless variety of motherboards and the many different combinations of SIMM and DIMM slots and memory types, it would be difficult if not impossible to assemble on a Web site all of the necessary information on how a particular memory error would map to a failing memory chip on a module.

Without the benefit of an actual memory testing device, or several on-hand memory modules to use for testing, trying to determine which module is bad is extremely difficult. There is, however, a tried and true method you can use to at least get started, and that is to swap modules around. Here’s a brief description:

Simple techniques for troubleshooting Memory failure without a Memory Tester.

Remove the modules one by one from motherboard
This is simplest method for isolating a failing module, but this may only apply if the motherboard has more than one memory module slot. By selectively removing modules one at a time from the system, and then test run the system, you will be able to find the bad module quickly. Make sure to mark each module as having either passed or failed.

Swap the modules around When none of the modules can be removed, swap and rotate modules to find which module is defective. This technique can only be used if there are two or more modules on the motherboard. If there are two modules, change the location of two modules.

As an example, place the module from slot 1 into slot 2 and place the module from slot 2 in slot 1, and then run the diagnostic software test again. If either the failing data bit or address changes, you know that one of the modules you have just swap is defective. By using several combinations of module swapping you should be able to check which module is defective.

Replacing with known good module
If you are unable to use either of the above two techniques, the only options left are to either contact a technician or use known good modules and selectively replace each of the modules one by one to pin point the memory failure. This is the easiest way to detect a memory failure.

Another Option is to remove and clean the metal contacts and motherboard slots
If your system is fairly old and/or in an area where there is allot of dust or tobacco smoke, sometimes the dust, nicotine or even oxidation will cause poor contact in the SIMM/DIMM slots. Remove the modules and clean the gold or tin contacts with a “pencil eraser” or any cleaning solution used for electronics such as video and audio head cleaning. In a pinch, a good grade of nail polish remover or acetone can be used for this purpose, just be careful and make sure to dry the modules and SIMM/DIMM slots thoroughly. Make sure to note which slot is a specific module came from, and be careful not to reverse modules when reinserting them into the SIMM/DIMM slot

Reminder, you can sometimes identify memory failures by using motherboards BIOS beep codes
This is especially useful if you are not trained to perform the correct diagnostic methods. All major motherboard manufacturers and BIOS developers have devised a simple way of informing you that your system is having a problem by emitting beep tones from the built in speaker on the motherboard when a problem arises.

Next we will focus on using diagnostic software to detect or confirm possible memory problems. One software product, DocMemory, a product of CST Inc – SIMMTESTER.com was developed to be a user friendly PC Memory Diagnostic Software tool. The diagnostic software can be installed on a single self-booting floppy disk, making it independent of any operating system.

Click here to Continue

Click here to go to the Performance Center Home Page