The old gaming PC restarts quickly but fails to post.
The old gaming PC restarts quickly but fails to post.
Hi everyone,
Details:
16 GB RAM (DDR3?)
Intel i5 processor
GeForce 970 graphics card
Asus Maximus Ranger VII board
I own a used gaming PC that was repurposed as a desktop for my fiancée. It's an older model and has been functioning well until recently. When I press the power button, certain lights begin to blink, all fans start spinning, but after 2-3 seconds the system powers off again and restarts repeatedly, without any post or beep sounds. The PC restarts on its own before it even reaches that point.
Short video (no sound) showing the problem:
https://imgur.com/a/XAu6ZPM
You can also check this link for more information:
https://imgur.com/a/XAu6ZPM
Here’s what I’ve tried so far:
- Reseat with RAM
- Reseat with graphics card
- Disconnect all drives
- Reset CMOS
- Replace the battery
- Installed a new power supply and tested it
- Disconnected power button and reset button, then booted using a screwdriver
I searched various forums for solutions but haven’t found any helpful answers yet.
Since I’m not very familiar with hardware, I’m unsure what steps to take next. The only option I considered before taking it to a repair shop was removing the PC from its case. However, I haven’t done that yet because I lack experience and would need to undo all cable connections and possibly remove several components, which seems risky and time-consuming.
I appreciate any advice or suggestions you can offer.
The present power supply is an advanced yet quite outdated Nesteq Xzero model, and I replaced it with a brand new BeQuiet System Power 9 to verify if it matches the PSU (700W each). However, the error still occurs in both instances, indicating the issue isn't tied to the power supply itself and the older one functions properly.
I have already attempted that. The same applies to RAM and drives. It's possible my description was unclear.
I've updated the subject to clarify what I tried.
Tomorrow I plan to take out the board from the case and put the computer together on the kitchen table.
This is usually related to the PSU, which is why I won’t go into detail. If the PSU is attempting to draw excessive power, the computer will show your issue. Another potential cause could be the CPU thermal paste. Most thermal paste needs to be reapplied annually, occasionally every four years, and Noctua NT-H2 lasts about five years. It’s recommended to remove the CPU heat sink and inspect the TIM (paste). If it appears dry or is cracked, it should be replaced. Also, TIM tends to thin over time, creating uncovered areas that lead to heat buildup.
There are other factors that might contribute, but checking the TIM is especially important for older systems. A faulty electrical connection—such as a motherboard touching the case, a bare wire contacting the case, or a malfunctioning fan—can also be responsible.
The thermalpaste is likely more than eight years old. Would an extremely damaged old thermal paste be enough to cause the PC to run for a few months before stopping completely because it finally failed? How long does it take for heat protection to activate once the system starts? Could that be the reason for the 1-2 seconds you notice right after powering on?
Yes, brief periods of ON followed by OFF can result from this.
The options are limited to a handful of CPUs.
First, I’d consider swapping the CPU if possible—any model that fits and works will suffice. This could be an i3 just to test whether the issue lies with the CPU or the motherboard.
It’s also worth trying to reinstall the CPU gently, ensuring it’s placed correctly in the socket. If you accidentally drop it, the pins might bend. Use a magnifying glass or a close-up photo to spot any misaligned contacts. Clean the old paste thoroughly and apply fresh paste.
I own many older LGA1150 boards, so swapping two processors should work fine. Otherwise, I’d look for a used CPU on eBay; most are reliable. A $10 to $15 upgrade could confirm if your i5 is faulty.
Thermal compound is only necessary when installing a new cooler or replacing the CPU. I’ve never encountered issues with long-lasting coolers.
For optimal heat transfer, use just a thin layer of paste—both too little and too much are ineffective. A pea-sized amount usually suffices for most CPUs (except possibly LGA1700 and larger Xeons). I might even experiment with graphene pads someday.
On older systems, I rely on a tube of white Aluminium Oxide paste from RS, which is over 25 years old. For newer models, Noctua NT-H2 is common. Arctic MX-4 may separate after a few months.
During startup, the BIOS searches for compatible CPU, RAM, GPU, and then looks for bootable drives. If your BIOS still works, it’s likely missing a key component—hence frequent restarts.
Have you tried booting with just one DIMM? Have you experimented with other DIMMs? If you have another compatible machine, swap the RAM and run MemTest86 from a bootable USB drive.
Right now, I’m leaning toward checking the CPU or motherboard.