Following the OC, there is no power up. When the CPU 4/8pin is not present on the motherboard, the system powers on.
Following the OC, there is no power up. When the CPU 4/8pin is not present on the motherboard, the system powers on.
I'm facing difficulties identifying which part is malfunctioning after a fairly aggressive CPU overclock.
The voltage was set to 1.603v, with the last setting before the motherboard BIOS displaying red warnings for unsafe voltages, and the CPU ran at a 225Mhz x18 ratio, totaling 4050Mhz.
During the stress test, after booting to the desktop and running monitoring tools, the system stayed stable with idle temperatures around 28°C.
A blend test on Prime95 showed full utilization across all cores for short durations, with temperatures staying below 55°C. Suddenly, power cut off abruptly, as if the connection was severed.
The problem seems to occur when unplugging the PSU, flipping I to O, holding the power button to discharge excess energy, then plugging back in and switching I again. The power button lights and fans briefly activate for under a second before shutting off completely. Repeating the process results in the same brief activation.
I've checked everything thoroughly—cleared CMOS, disassembled the PC (including the CMOS battery), left it overnight, tested various component combinations, and still encountered the same brief flicker issue.
Additional steps included disconnecting the PSU from the motherboard, changing the power connector from green to black on the 24-pin port, which allowed fans and drives to run. After reassembly, the system booted normally with all lights on, but no CPU was present.
It appears testing with a different PSU will clarify whether the issue lies with the power supply.
If it starts up and I can access the BIOS using another PSU, the original unit is likely faulty (especially the 4/8 pin CPU connector).
If it fails to start with any PSU, the motherboard is probably the cause.
Anyone have insights on which component might be the culprit before purchasing a new PSU and motherboard?
JulienJiJiMuGe :
I'm facing issues figuring out which part isn't functioning properly after a minor CPU overclock.
The OC:
CPU voltage is set to 1.603v (the last step before the motherboard bios displays red text for dangerous voltages) and the CPU ratio is 225Mhz x18, totaling 4050Mhz.
"slightly aggressive CPU overclock", ROFL :lol:
To be honest, your CPU voltage switches to Red Text at 1.425v. I assumed that wasn't right, so I verified it on my own MSI M/B.
1.425v is also the maximum design voltage for your 1090T. Even though you were close to the red danger zone, you chose to run it at that level.
Do you even know how many voltage settings exist between 1.425v and...?
Another reason I don't favor recommending the hyper 212 evo for heavy overclocking is that at such high voltage levels, your CPU could exceed 80°C temperatures, which is quite risky. It seems to be limited by the CPU itself. Try pressing the CMOS reset button on your motherboard and then reset the overclock settings.
Another reason I avoid recommending the hyper 212 evo for heavy overclocking is that at such high voltage your CPU could exceed 80°C temperatures, which seems quite risky. It appears to be a CPU-limited situation. Try pressing the CMOS reset button on your motherboard and reset the overclock settings.
As for your diagnosis, did you read his ten-paragraph analysis? I’m pretty sure he did that during the first 20 minutes when it hit the limit.
For my side, I’ve dealt with PSU issues and it aligns with your description about the fans lighting briefly. However, I managed to boot up but without the CPU power connectors, which makes the PSU failure less likely. Still, I’m not saying it’s a perfect failure—just a weak PSU that didn’t last well.
I’m curious what you did before you went for APSHUT and adjusted the voltage to 1.6V on the core. Were you just experimenting and this was the problem you couldn’t fix?
*edit before you buy more parts... consider getting an Intel chip. I did that and it really pushed my non-K i5 from 3.3GHz to 4.5GHz. I can tell you exactly what changes to make while I’m still here. Skylake is straightforward as long as you have the right hardware.*
JulienJiJiMuGe :
I’m facing some issues figuring out which part isn’t functioning properly after a bit of aggressive CPU overclocking.
The OC:
CPU voltage set to 1.603v (the last step before the motherboard bios displays red text for dangerous voltages) and a 225Mhz x18 CPU ratio, totaling 4050Mhz on the CPU.
"slightly aggressive CPU overclock", ROFL :lol:
To be honest, your CPU voltage switches to Red Text at 1.425v. I assumed that wasn’t right, so I verified it myself using my MSI M/B.
1.425v is also the maximum design voltage for your 1090T. Even though you were well into the red danger zone, using 1.603v still seemed risky.
Do you really know how many voltage levels exist between 1.425v and 1.603v? It probably wouldn’t have worked at 1.500v with a 200Mhz X 21x setup (4.2ghz).
But now you’re wondering what settings fall in between—unless you find another 1090T on Ebay, you might not know.
Well done, you’ve damaged your CPU!
Edit: I saved this thread as an example of what not to do with CPU voltage beyond specifications. It also highlights the importance for overclockers to understand their hardware limits before attempting changes.
For clarity, there was no need to increase the base clock from 200mhz to 225mhz on the unlocked multiplier 1090T. Doing both would have caused instability, and then adding a cherry on top—running 1600Mhz memory speed on a 1333Mhz CPU—was the final twist. It’s still alive!
Can I ask you something? Have you learned anything from this?
Your problem matches what I experienced. You used the same CPU cooler but changed the fan to a Silverlight adjustable and added a Magma unit on the other side. Recently, I adjusted my system slightly. Running 3DMark showed temperatures staying between 45-50°C, which caused it to shut down. It won’t boot into POST even after trying all fixes. MSI helped me understand that it took two weeks for a refurbished board replacement, so I prefer keeping it sealed and never opening it.
@alexoiu - I'm going to finally follow your suggestions and try it with a different CPU next. 😊
@Chugalug_ - I understand that the 212 EVO isn't a top cooling option, but it managed to keep the CPU at its maximum of 55 under load and in the high 20s during idle. However, as you mentioned, the excessive voltage was the real cause of failure.
@L0stChild - For your interest, I previously attempted overclocking several years ago, achieving a stable 3.6Ghz with the turbo core. When upgrading to a new GPU, I wanted more headroom for the aging 3.6Ghz speed, so I looked into what others had done with their 1090t and realized that accepting a 1.6v was essentially a dead end.
@4Ryan6 - It definitely failed! I had been reading through that information and it only displayed as white.
🙁
In any case, it was clearly operating near the dangerous limits...haha! You're right about using the maximum design voltage as a reference. Even though I've successfully overclocked before, this experience taught me a valuable lesson.
@eightydee - I also replaced my motherboard, and while it helped somewhat, it didn't fully resolve the problem. In that instance, I resorted to overvolting the CPU.
I realize that pushing 1.6v was far beyond what was reasonable (understatement), and I thought it might be a valid target after reading other discussions.
The conclusion was that rushing to accept an acceptable voltage was a mistake.
Just for transparency, since I've tested various components with working ones, except the CPU:
- The PSU performed well on another PC.
- After installing a new motherboard, it would boot but would experience random freezes, sometimes restarting after 5 seconds or even shutting down completely—this happened during POST, BIOS, and Windows. Occasionally lasting up to 15 minutes.
- I experimented with different RAM, but no improvement was noticed.
- I also tried various configurations: starting with minimal components, using a different PSU, removing one of the two 4-pin CPU connectors (which was the only time it lasted 15 minutes before freezing or powering off), applying multiple underclocks and BIOS tweaks to try and stabilize it. None worked.
- My next step is to install a new CPU. I'm considering the FX 6300, which fits well in the AM3+ socket on the ASRock Fatality 970 Performance board I acquired.
In summary:
Take @4Ryan6's advice—identify the manufacturer's maximum design voltage rating, not just random anecdotes from overclocking discussions. You can exceed it, but do so gradually and carefully. Unless you have spare parts and a good budget, proceed with caution.
And as I've always known (and felt confident enough to try again), the key to successful overclocking is starting small, making incremental changes, and slowly increasing settings.
It seems it was your CPU that caused the problem. I need to check the motherboard because I'm sending it back to Newegg unless it's the CPU. I can RMA the CPU, but I only have a few more days. I changed the motherboard and it started working again, which suggests the old one was damaged too. Now I also need to replace the CPU since it keeps shutting off and freezing randomly. From your first message: I thought about resetting the CMOS before returning. If you haven't done that yet, it might be the same issue as with both the motherboard and CPU. I can't guarantee it's the same problem because you only did a little tweaking, and I tried overclocking at 1.6v, which was quite intense.