Hardware Diagnostics

General

Blow dust out of machine before starting any diagnostics. When blowing dust out of a fan (including case fans, CPU fans, and fans inside the power supply) it is important that you prevent or minimize the fan’s spinning. Using compressed air may rotate the fan much faster than its rated limit and the sudden burst of air pressure can damage the fan blades, especially for smaller and/or older fans where the plastic may be thin or brittle.
    For exposed fans like those on desktop CPU, GPU, and case fans it is sufficient to hold the fan in place by pressing gently on the center of the fan to prevent it from spinning.
    For fans that are covered or obstructed, such as the CPU fan in a laptop, use a small straightened paperclip to block the spinning motion of the fan by inserting the paperclip in the gap between two fan blades. Do not use excessive force to do this and be careful not to damage any grates or vents on the exterior of the case.
    Do not stick any objects inside a power supply to prevent the fan from spinning! Power supplies carry enough voltage to be dangerous and you can harm yourself or the customer’s computer in the process.
    For power supply fans and other fans you cannot safely prevent from spinning, use lighter bursts of air and let the fan completely stop spinning between bursts to minimize the amount the fan spins.
Check device warranty with manufacturer. Some repairs are more reasonably handled as an on-site or mail-in repair with the manufacturer under warranty if available, and may save everyone involved time and money. Most manufacturers have an easy-to-find warranty check tool (Acer, Asus, HP, Dell, Apple, and Lenovo, to name a few). For example, a quick Google search for “HP warranty check” will return the HP Warranty Lookup tool.
    Copy and paste the warranty information into a tech note on the device’s work order.
    You can also determine a bit more about the device from the warranty page, usually. Specifically, the purchase date and/or warranty expiration date may let you know that the device is too old for a cost effective repair, or new enough that the customer could extend their warranty to have the device repaired by the manufacturer.
Check for exterior damage to the machine. This step should have been briefly performed with the customer already during the check-in procedure, but use this time to perform a more thorough inspection. If the damage is visible, take a picture and attach it to the WO for record keeping.
    Look for cracks, missing or broken components, evidence of liquid damage, excessive dirt and grime, burn marks, dust, tobacco smells, etc.
    Bugs- if you find bugs or evidence of bugs (roaches, ants, etc), turn the device off and bag it up in a trash bag. Seal the trash bag and wrap that in another trash bag, and seal it. Call the customer and inform them that the device is a health hazard and that they will need to address that issue on their own time before bringing the device back to us.
If the customer has complained of a specific issue (such as overheating, or a loose component, device not charging, etc), address this now.
If the device requires disassembly to test specific hardware issues, and a disassembly fee has been approved, do so now.

PC

Run extensive/full built-in diagnostics when available. The following manufacturers include built-in diagnostics for most models:
    Dell
    HP
    Lenovo (run from CD)
For all others, use either Hiren’s Boot CD or Ultimate Boot CD and run:
    WD Data Lifeguard (actual HDD model doesn’t matter, this will test any drive)
        If WD Data Lifeguard doesn’t recognize the drive, remove the drive and hook it up to the shop Linux computer. Use the information in the Linux section below to troubleshoot the hard drive further.
    Memtest86+
When finished, take note of success and/or failures on the tests. Add Hard Drive Pass/Fail and Memory Pass/Fail actions to the WO as appropriate.
If all tests passed, boot the computer into the operating system and proceed with Software Diagnostics.
If memory failed, isolate which stick(s) or slot(s) are failing and assess repair options.
If the hard drive failed, attempt to back up the device on the shop server. See Procedure - PC Backup for more details.

Mac

Check to see if any damage is on the outside of the machine. If found, take pictures and attach to work order.
    Damage includes “dog-eared” corners of the device, dents that are easily noticeable several feet away from the device, heavy gouges in the case, screen damage including cracked or shattered glass panel or damaged LCD panel underneath, damage to the keyboard or trackpad, liquid damage, loose components, etc.
Disassemble the device according to repair articles in GSX to check for physical or liquid damage, burn marks, dust buildup, etc.
Boot the computer to AST or AST2. For Macs released before 2015, hold down the N key while powering on the machine. For devices released in 2015 and later, press and hold Option+D while powering on the machine. In both cases you will be taken to the AST menu. Release the held key(s) once the AST menu appears.
Run MRI (Mac Resource Inspection) and make a tech note in the work order with the results.
Restart machine and hold down the same key(s) as before. This time, choose the option to boot into Tech Tools. Select the newest version of Tech Tools that can be used on this computer.
Launch Tech Tools from the dock. Run the disk drive surface scan, SMART scan, video, and memory tests. Make a tech note with the results.
Launch Disk Utility from the Utilities folder in the Applications folder. Verify or repair the disk, depending on what options are presented in Disk Utility. Options vary by version.
Shutdown the computer and boot onto the primary boot disk.

Linux Hard Drive

Find the device identifier using the Disks program. The disk will be listed in sequential order by the date/time they were first recognized by the system (so the drive you are looking for will usually be toward the bottom of the list). The disk identifier looks like /dev/sdX, where X is a letter a-z (such as /dev/sdb or /dev/sdc).
Open a terminal window and use command sudo badblocks -v /dev/sdX > /media/pclogs/#####.log, where /dev/sdX is the disk identifier from the previous step and #####is the PCID of the computer in question.
    If you are backing up PCID 12345 and the disk identifier is /dev/sdg, then the command will be sudo badblocks -v /dev/sdg > /media/pclogs/12345.log.
    The program will check for bad blocks on the device and alert if there are any present. If the scan completes with no errors then there are no bad blocks present. A successful scan will end with the following: Pass completed, 0 bad blocks found. (0/0/0 errors). Anything else constitutes a drive failure.
Additionally, run GSMARTControl to retrieve the SMART information from the drive.
Use the PasteCare utility (https://pcrt.care/paste) to upload to PCRT the terminal output of the badblocks command and the raw SMART report from GSMARTControl.