This chapter is from the book
This chapter is from the book
This chapter is from the book
Troubleshooting RAM
It’s not common, but RAM modules can cause intermittent issues, or they can fail altogether. Always make sure that the RAM is fully seated within the RAM slot and that the plastic ears are locking the RAM into place. An unstable system can be caused by several components, including RAM. Remember to check other components in the system as well, including the video card, motherboard, and CPU.
A lot of the issues you see occur because a user has purchased and installed a memory stick that is not compatible, or is semi-compatible, with the motherboard: wrong speed, incorrect capacity, improper configuration, and so on. Be ready for this; check the RAM compatibility against the motherboard, even if the user swears it has been checked already. Remember, a good technician has documentation available, has access to the Internet, and knows how to use both.
Perhaps there was some kind of surge inside the computer; maybe the computer is not protected by a surge suppressor or UPS. Another possibility is that the RAM was damaged by ESD, and this damage manifests in intermittent problems. There are expensive hardware-based RAM testers that can tell you if the RAM is electrically sound and if it can process data correctly. If your company owns one, or if you can get your hands on one for a short time, you might narrow the problem. However, we have rarely needed to use these.
Here are some possible symptoms of a RAM issue and corresponding troubleshooting techniques:
Computer will not boot/or randomly shuts down: If there is no RAM in the computer, or if the RAM is damaged or is not installed securely, it can prevent the computer from doing anything at all (aside from draining electricity from your AC outlet). For example, the power supply fan turns but nothing else happens—no beeps and no displays. First, if the RAM was just installed, make sure that the RAM is compatible. Next, and in general, try reseating the RAM before you attempt to troubleshoot a CPU or motherboard. Add RAM if none exists. (Sounds silly, but we’ve seen this!) If you suspect faulty RAM, corroded contacts, or a faulty RAM slot, you can try taking the RAM out, cleaning the RAM and RAM slot, and putting the RAM back in, being sure to seat the memory module properly. (For cleaning, use contact cleaner on the RAM contacts and use compressed air on the slot.) Next, try moving memory modules to different slots; check your motherboard documentation for proper orientation. As mentioned, a POST analysis can be helpful in these situations as well. If necessary, replace the memory module with an identical one (if you have an extra one handy) or purchase a new one if you have identified the memory module as the source of the problem. In some cases, RAM can overheat and cause random shutdowns. To prevent this, purchase a heat sink for RAM. These are made of aluminum or copper, just like CPU heat sinks, and are sometimes referred to as heat spreaders. You can also purchase RAM with heat sinks preinstalled. This type of RAM might be necessary for high-end systems.
BIOS indicates a memory error: The BIOS can indicate a memory error through a message on the screen and a flashing cursor or by beeping. If it beeps, you need to reference the motherboard documentation for the specific beep codes. Sometimes a BIOS setting can be incorrect. If the computer has a saved version of the BIOS settings, you can try reverting to them, or you can try loading the BIOS defaults; we can’t tell you how many times this has worked for me! Sometimes the BIOS indicates the wrong amount of RAM. If this is the case, check the RAM, as explained in the first bullet. Finally, a BIOS update can be the cure; perhaps the BIOS just doesn’t have the programming necessary to identify the latest type of RAM that was installed.
An application crashes: If an application crashes, it could be due to faulty RAM (or, uncommonly, a CPU issue). Make sure that the RAM is compatible with the motherboard and use memory testing programs to verify if there is any damage to the RAM. (More on memory testing in just a bit.)
A stop error, aka BSOD (blue screen of death), occurs: This is a critical system error that causes the operating system to shut down. Most of the time, these errors are due to device driver errors (poor code), but they can be associated with physical faults in memory. One example of this would be a non-maskable interrupt (NMI). An NMI can interrupt the processor to gain its attention regarding nonrecoverable hardware errors, resulting in a BSOD. The BSOD usually dumps the contents of memory to a file (for later analysis) and restarts the computer. If you don’t encounter another BSOD, it’s probably not much to worry about. But if the BSOD happens repeatedly, you should write down the information you see on the screen and cross-reference it to the Microsoft Support website, at https://support.microsoft.com. Again, if you suspect faulty RAM, try the troubleshooting methods in the first bullet.
Chances are you won’t need them often, but memory testing programs such as MemTest86 (https://www.memtest86.com) are available online. In fact, a whole slew of testing and benchmarking software is freely available to you. Search around! Plus, you can use the Windows Memory Diagnostics tool, which you can access by typing mdsched.exe in the Run prompt or from the Windows Recovery Environment (more on that in Chapter 53). These steps can help diagnose whether a memory module needs to be replaced. But in general, trust in your senses. Look at and listen to the computer as you try to diagnose any RAM issues that might occur.
