CPU and memory indicators flash regularly.
CPU and memory indicators flash regularly.
My setup has stopped working after a decade of use. I had already arranged to swap out parts, so it wasn’t a surprise. However, the situation worsened because I still lacked funds for a new system. I’m trying to keep my PC running until it can be replaced soon. Any guidance would be greatly appreciated.
My problem: I recently changed from an EVGA GTX 960 SSC to a Gigabyte RTX 3070 Gaming OC. At the same time, I upgraded my power supply from a Seasonic S12II 520W to a quiet Power 11 750W unit. Initially, everything functioned well—playing old games with maximum graphics settings. But after a few days, the PC would suddenly stop. It happened three times while using Google Chrome (only that browser). The second time it froze completely and resumed afterward. The first and third instances required a hard restart by holding the power button. After that, I considered updating the BIOS, but it wasn’t available. Now, when I power on the system, the CPU LED stays illuminated for about half a second, then switches to DRAM LED, and the CPU LED flickers back briefly before the cycle repeats. No startup occurs.
I attempted to update the BIOS using the built-in flashback button, but it failed. The USB drive is formatted as FAT32, and the .cap file was renamed to Z77ST.cap. Still, the blue light from the flashback button remains on for minutes even when power is supplied but the system is off. I’ve tried this with all components disconnected—adding just one RAM stick, none at all, keeping the old PSU, graphics card, and so on—while pressing the MEM_OK button.
Sometimes after a failed POST, the CPU LED stays lit. This suggests the processor might be faulty. I’ve ordered a used i5-2500K to test, but since I can’t upgrade my BIOS, it won’t help much. Do anyone know about dead CPUs from this era or this board? My setup includes:
- Intel i7-3770K (never overclocked)
- Asus Sabertooth Z77
- Four 4GB G.Skill DDR3-1866 RAM modules
Have you attempted just resetting the CMOS? Swapped in the processor again? Verified the connections at the CPU socket?
I've worked with older systems quite often, though never this exact model. The pattern of flashing debug lights usually points to a memory problem. It could be a misconfigured BIOS setting that prevents proper memory initialization, a faulty RAM stick, a defective CPU memory controller (unlikely unless you were stressing the RAM too much), sometimes a dead CMOS battery causing the issue, or in rare cases, the board simply failing after a decade—about the typical lifespan for such hardware. A corrupted BIOS is definitely a possibility, but it wouldn't be my first suspicion. I'd be surprised if your CPU was completely dead; they're quite resilient unless you're pushing them beyond safe limits during overclocking. Usually, when a CPU fails without overclocking, it's because another component on the board malfunctioned and took it with it.
First of all, thank you for your quick replies! Unfortunately I just found out that I can reset the CMOS RTC RAM-Data by applying a Jumper for 10 secs. I'll definitely try this out as soon as I'm home again. At the same time I don't think that has to do with an incorrect placement of the CPU since it's been running now for 10 years without any issue whatsoever. But I will def. check that as well, when I'm exchanging the CPU. To be entirely honest, I doubt that there's an issue with the DRAM, since they were working fine minutes before the last shutdown, as they were correctly displayed in the device-manager. Furthermore I've pressed the MemOK!-Button several times before and during POST, which "ensures memory boot compatibility. [...] MemOK! determines fail-safe settings and dramatically improves your system boot success" (quote: User manual of the MB). Also theres no difference in its behaviour when having all, just one or none sticks installed. The option of a dead CMOS-Battery is a possiblity since this occured after having exchanged a PSU after years of usage to a more powerful one (please correct me if im wrong) I agree with your assumption that the MB may simply be dead. As just mentioned, I think that having a constant current for 10 years and then an upgrade to a nearly 50% more powerful PSU, simply fried the board. But I still hope that this isn't the case. But could that be? Is it possible the new more powerful PSU burned the system after years of being used to a lesser powerful one? But now I'm worried, the MB may also destroy the new-ish CPU I've ordered. Could it may NOT be advisable to exchange it? Also, I would like to know, if the CPU has to be installed and in a working condition in order to flash the BIOS via the Flashback-Button on the backside of the MB.
Memory can vanish unexpectedly. In my experience, the MemOK! feature (which I've used with many ASUS boards) often fails, especially if the RAM has failed. It doesn't function well when the power supply just stops working. The real issue seems to be a small battery—typically a CR2032—that's quite old now. This problem likely occurs regardless of whether you replaced the PSU. Each board reacts differently: some lose BIOS settings on every startup, others show a CMOS warning screen, some need you to reset CMOS before turning on, and a few refuse to boot unless they have a standard 3V CMOS battery. Your understanding of how a PSU operates might be off. It supplies power rather than forcing it. A PSU upgrade simply means the board can handle more power, not that it will draw more. If you're using the same components, the same power output remains. The new PSU must be designed to handle higher demands—otherwise, performance won't improve. If the back panel has a button, it's usually not functional. I'll verify the manual or check the 24-pin and EPS 8 pins. If you need a working CPU and RAM, the BIOS flashback might not be worth it. However, if you're concerned, use a multimeter to test the VRM output on the board. It's easier to remove the plastic cover for a closer inspection. You can also measure resistance from an inductor to ground using a capacitor or one of the 8-pin pins. Ensure no CPU is present and check each inductor. If resistance is under 1-2 ohms, you likely have a ground short on a VRM—making the board unusable. A higher reading suggests checking the 12V power path from the EPS connector to the inductors. A very low value means the CPU could be destroyed when power is applied.
Initially, thank you for your thorough response. I would definitely consider it if tools were available. For now, I gave it another shot using a refurbished ga-z77p-d3. Being simple, I rebuilt my whole setup to be ready and hoped it would boot up properly. Unfortunately, there was no signal from the HDMI port or the GPU. The system just kept running but in a standby state. The same happened with the i5, double RAM, single RAM, no GPU, no memory, and no CPU. I’m planning to upgrade to a more suitable budget gaming rig featuring a Ryzen 7 5800X, paired with two 16GB Corsair CL18 dimms at 3600MHz, along with a Gigabyte X570S UD. This should work well with my existing GPU and software. I’m hoping it will last for several years; good luck! Still, I’m concerned about the second-hand RTX 3070. It did handle FH4 (with HDR active) and COD:BO3 at its maximum of 4k without any problems. I really think there’s nothing wrong with the BIOS chip. It should start up normally on Power 11 750W. Could it be that I accidentally damaged one or both components during this process? Or could it pose a risk to the parts I’m planning to install? Thanks again!