================================================================================
  HP ALL-IN-ONE  -  FREEZE / SPONTANEOUS REBOOT  -  DIAGNOSTIC KIT
  CoolTech Computers  -  Colin Clohesy
================================================================================

  THE COMPLAINT
    - Works for an hour or two, then freezes completely
    - No keyboard response, no mouse response
    - Occasionally restarts itself instead of freezing
    - Does NOT go to sleep

  WHAT'S IN THIS FOLDER

    Run-Diagnosis.cmd          <- DOUBLE-CLICK THIS FIRST
        Runs Get-HPAIODiagnosis.ps1. Read-only. 30-60 seconds.
        Writes a REPORT.txt to the Desktop.

    Run-Live-Monitor.cmd       <- USE THIS IF IT HASN'T FROZEN YET
        Runs Watch-HPAIO.ps1. Samples every 30 seconds to a CSV.
        Leave it open and use the machine normally. When it freezes,
        the CSV on disk shows what was happening in the minutes before.

    Get-HPAIODiagnosis.ps1     the history/evidence collector
    Watch-HPAIO.ps1            the live monitor
    README.txt                 this file


================================================================================
  HOW TO USE IT (over remote access)
================================================================================

   1. Copy this whole folder onto the machine
   2. Double-click  Run-Diagnosis.cmd
   3. Click Yes at the UAC prompt
   4. Wait 30-60 seconds
   5. It writes a folder to the Desktop called  HP-AIO-DIAG-<date-time>
   6. Send that whole folder back

   THEN, if the machine has not frozen yet:
   7. Double-click  Run-Live-Monitor.cmd
   8. Leave it open, use the machine normally until it freezes
   9. The CSV on the Desktop has the evidence


================================================================================
  THE KEY INSIGHT - READ THIS BEFORE LOOKING AT THE REPORT
================================================================================

  The very first section answers the question that splits the whole diagnosis:

      DID THE MACHINE CRASH, OR DID IT JUST DIE?

    A CRASH (BSOD) means Windows had time to write a bugcheck before it went
    down. That points at software, driver, RAM, or thermal - NOT a dead PSU.

    A POWER COLLAPSE means the machine simply stopped, leaving no complaint.
    That points at the power path or a thermal cutout.

    Both need completely different repairs, so the script settles it first.

  THREE WAYS THIS PRESENTS IN THE LOG:

    Bugcheck present (Event 1001/1003)  ->  IT CRASHED. Read the code.
    WHEA present (Event 18/19/47)       ->  HARDWARE error. Named component.
    Both empty + Event 41               ->  IT LOST POWER. No time to complain.

  IMPORTANT - EVENT 41 TIMESTAMPS LIE
    Event 41 records when the machine CAME BACK, not when it died. The entry
    is written at the next boot, because that is the first moment there is a
    running OS to write it.

    Event 6008 is the one that NAMES the actual moment of loss. Use 6008 for
    the timeline.

    And if the customer says "it only happens when I turn it on" - they may be
    right about what they SEE while the log genuinely cannot say when the
    power went. Do not correct them; explain the artifact.


================================================================================
  WHY AN ALL-IN-ONE IS THE WORST CASE FOR COOLING
================================================================================

  Everything is in the screen housing. The fan pulls air straight through the
  heatsink, and a domestic room will clog it. This is the single most common
  cause of "fine for an hour, then it hangs".

  The signature is unmistakable:
     - cold machine works perfectly
     - after 1-2 hours of heat soak the heatsink can no longer dump the load
     - the CPU throttles hard, or the machine cuts out

  IT IS THE SAME SYMPTOM WHETHER IT IS DUST, A DRY FAN BEARING, OR AGED
  THERMAL PASTE. The software can tell you it was hot. Only a screwdriver
  and a can of air will tell you why.


================================================================================
  WHAT THE REPORT COVERS
================================================================================

   1. CRASH or POWER LOSS     the deciding question, plus full shutdown history
   2. THERMAL                 live temps, thermal events, fan, throttling
   3. POWER                   shutdown type, sleep events, power config
   4. HARDWARE HEALTH         disk SMART, disk errors, memory, DIMMs
   5. DRIVERS AND DEVICES     problem devices, GPU driver, TDR events
   6. RECENT CHANGES          updates, drivers, apps - did it start after one?
   7. LIVE MONITOR            instructions and a snapshot
   SUMMARY                    verdict, counts, and what to do next


================================================================================
  IS IT THE SSD?
================================================================================

  Worth ruling out early, and it is FREE to rule out - the log does it.

  Watch for Event 129 "Reset to device was issued" - a storage controller
  reset. That is a known cause of a hard freeze with NO bugcheck, because
  the machine waits on the disk that has stopped answering.

  Also check:
     - Get-PhysicalDisk HealthStatus  (should be Healthy)
     - the SMART counters in section 4a
     - Event 7 / 51 / 55 in section 4b

  A failing drive and a failing PSU produce the SAME symptom: intermittent
  hard stop with no warning. The log tells them apart. A drive fault leaves
  storage events. A power fault leaves nothing at all.


================================================================================
  HONEST LIMIT
================================================================================

  Host-level logs can prove THAT it crashed, lost power, or overheated. They
  cannot see a dry joint, a failing fan bearing, a swollen capacitor, or dust
  you would find in thirty seconds with the cover off.

  When the log says "power" and the machine is old, the next step is a meter
  and a screwdriver - not more software.

================================================================================
