Bsod after only increasing vcore?

[Ace McPlane]

I’m getting a blue screen of death after only increasing the vcore a bit. It stress tests for over 5 hours fine at 1.235v but crashes at 1.24 with nothing else changed.

 

temperatures are good, multiplier is 47, ring ratio is 42, vring is under vcore, vccin is around 1.9 max.

 

this is an i7 4790k on an msi z97 pc mate which doesn’t have a lot of bios options even after updating. No control over llc for example and turbo boost automatically turns off with EIST.

 

1.24 vcore is actually 1.264v max due to vboost.

 

It doesn’t make sense to me that increasing voltage would cause a bsod when it worked fine with less. Usually it’s the opposite where increasing voltage stops the bsod.

 

any thoughts?

[porina]

What is the actual temperature? "good" doesn't cut it. I've had systems that go unstable on hot days for example, at temperatures still generally considered safe. 70's seem generally ok, 80's when stability starts to wobble.

[paddy-stone]

Could be that the 1.235v's stress tests gave you a false sense that it was good maybe?  I can't see any other reason why increasing the voltage gave you BSODs. I'd try increasing the voltage again personally, you're nowhere near max vcore. My i7 6700K needed around 1.28vcore for 4.7Ghz at the best I could whittle it down IIRC.

[Ace McPlane]

2 hours ago, porina said:

What is the actual temperature? "good" doesn't cut it. I've had systems that go unstable on hot days for example, at temperatures still generally considered safe. 70's seem generally ok, 80's when stability starts to wobble.

73 degrees

[Ace McPlane]

2 hours ago, paddy-stone said:

Could be that the 1.235v's stress tests gave you a false sense that it was good maybe?  I can't see any other reason why increasing the voltage gave you BSODs. I'd try increasing the voltage again personally, you're nowhere near max vcore. My i7 6700K needed around 1.28vcore for 4.7Ghz at the best I could whittle it down IIRC.

Yea I don’t get it.

maybe I need to bring my ring ratio and vring down to increase the voltage? Not sure. I’ve done a ton of stress testing at 1.235 with no issues so it just doesn’t make any sense. A 1.9 vccin should be more than enough for the voltage I’m drawing as well.