boriss911

Commendable
Nov 1, 2019
68
8
1,535
Hello, i am using ml240lrgb in combo with i7-3770k that was oc to 4.6ghz. Had that in about a year, everything was working fine, 35 idle, 80 gaming. Recently i was using speedfan to make my fans quieter. After i took bencmark, i saw that cpu went up to 100 degrees. Went back to speedfan, turned on fans to 80%, but cpu was still hot. I dont get it, fans are turned back on, but cpu temps wont go down. I am going to remount aio, and change thermal paste.
 

rodrigoxm49

Great
Oct 13, 2019
119
8
95
at 4.6ghz 1.35v, 4.2ghz 1.2v, it was stable even at 4.6ghz before, but now its even hot at 4.2ghz
1.35 is not that much. A AIO should be more than enough to keep it below 75C even on Prime. Make sure that there's no problem with AIO assembly. Sometimes even a little mistake can make temps to too high.
 

boriss911

Commendable
Nov 1, 2019
68
8
1,535
1.35 is not that much. A AIO should be more than enough to keep it below 75C even on Prime. Make sure that there's no problem with AIO assembly. Sometimes even a little mistake can make temps to too high.
Thats why i am going to change thermal paste etc, but i will try to turn speedfan off too, cause that could be the problem too. But i hope that AIO and CPU is fine.
 

CompuTronix

Intel Master
Moderator
... ml240lrgb ... i7-3770k ... oc to 4.6ghz ... about a year ... 35 idle, 80 gaming. Recently ... took bencmark ... 100 degrees ... going to remount aio, and change thermal paste.
Which "benchmark"?
At what ambient temperature?

at 4.6ghz 1.35v, 4.2ghz 1.2v, it was stable even at 4.6ghz before, but now its even hot at 4.2ghz
Your AIO may be prematurely failing.

1.35 is not that much. A AIO should be more than enough to keep it below 75C even on Prime ...
Prime95 v29.8 Small FFT's, all AVX selections disabled.

Guys,

Respectfully, you've both apparently been misinformed concerning Core voltage. Processors of different microarchitectures vary in voltage tolerances, which drives thermal behavior, as well as longevity. Regardless of what you may have "read" or "heard", it's likely that you've been misled into thinking that "1.35 is not that much". While this is true for 14 nanometer processors, 1.35 is indeed too much for 22 nanometer processors. Accordingly, a "one size fits all" approach does not apply to maximum recommended Core voltage.

No two processors are identical; each is unique in voltage tolerance, thermal behavior and overclocking potential. Overclocking is always limited by two factors; voltage and temperature.

We know that over time, excessive voltage and temperature can damage electronics. In CPUs, this can result in accelerated "Electromigration" - https://www.google.com/?gws_rd=ssl#q=Electromigration

This prematurely erodes the traces and junctions within the processor's layers and nano-circuits, which will eventually result in blue-screen crashes that become increasingly frequent over time. As a rule, CPU's are more susceptible to Electromigration or "Degradation" with each Die-shrink. However, the most notable exception is Intel's 14 nanometer Microarchitecture, where advances in FinFET transistor technology have improved voltage tolerance, which does not include 22 nanometer 3rd Generation processors such as the i7-3770K.

Here's how the Degradation Curve on 14 nanometer 5th through 9th Generation differs from 22 nanometer 3rd and 4th Generation:

iQuLSzu.jpg


Degradation Curves are relative to the term “Vt (Voltage threshold) Shift” which is expressed in millivolts (mv). Users can not monitor Vt Shift, which basically represents the potential for permanent loss of normal transistor performance. With respect to overclocking, excessively high Core voltage drives excessively high Power consumption resulting in higher Core temperatures during heavy workloads, all of which contribute to Vt Shift over time. Core voltages that impose high Vt Shift values are not recommended.

Here's the maximum recommended Core voltage per Microarchitecture from 14 to 65 nanometers since 2006:



For 22 nanometer processors, 1.35 Vcore is above the Degradation (Electromigration) curve. Even when running big air or 240mm or larger AIOs, at 100% workload overclocked 22 nanometer processors become increasingly difficult to cool as Vcore approaches 1.300. This is because unlike 2nd Generation and earlier processors which have a soldered IHS, 3rd through 8th Generation processors instead have TIM between the IHS and the Die. This is why "delidding" became a solution for overclockers.

Remember to keep overclocking in perspective. For example, the difference between 4.5 GHz and 4.6 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.

Although "Throttle" temperature for the i7-3770K is 105°C, it’s not advisable to run your CPU near its thermal limit. The consensus among well informed and highly experienced system builders, reviewers and overclockers, is that cooler is better for ultimate stability, performance and longevity. Experts all agree that it's prudent to observe a reasonable thermal margin below Throttle temperature, so here's the nominal operating range for Core temperature:

Core temperatures above 85°C are not recommended.

Core temperatures below 80°C are ideal.



Keep in mind that Core temperatures increase and decrease with ambient (room) temperature, for which the international "standard" for "normal" room temperature is 22°C or 72°F.

I suggest that you consider setting your Vcore to the maximum recommended value of 1.300 and configuring your highest stable overclock accordingly.

We have a "Sticky" that covers this topic and is recommended reading: Intel Temperature Guide.

CT :sol:
 
  • Like
Reactions: toniplavna

boriss911

Commendable
Nov 1, 2019
68
8
1,535
Which "benchmark"?
At what ambient temperature?


Your AIO may be prematurely failing.


Prime95 v29.8 Small FFT's, all AVX selections disabled.

Guys,

Respectfully, you've both apparently been misinformed concerning Core voltage. Processors of different microarchitectures vary in voltage tolerances, which drives thermal behavior, as well as longevity. Regardless of what you may have "read" or "heard", it's likely that you've been misled into thinking that "1.35 is not that much". While this is true for 14 nanometer processors, 1.35 is indeed too much for 22 nanometer processors. Accordingly, a "one size fits all" approach does not apply to maximum recommended Core voltage.

No two processors are identical; each is unique in voltage tolerance, thermal behavior and overclocking potential. Overclocking is always limited by two factors; voltage and temperature.

We know that over time, excessive voltage and temperature can damage electronics. In CPUs, this can result in accelerated "Electromigration" - https://www.google.com/?gws_rd=ssl#q=Electromigration

This prematurely erodes the traces and junctions within the processor's layers and nano-circuits, which will eventually result in blue-screen crashes that become increasingly frequent over time. As a rule, CPU's are more susceptible to Electromigration with each Die-shrink. However, the most notable exception is Intel's 14 nanometer Microarchitecture, where advances in FinFET transistor technology have improved voltage tolerance, which does not include 22 nanometer 3rd Generation processors such as the i7-3770K.

Here's the maximum recommended Core voltage per Microarchitecture from 14 to 65 nanometers since 2006:



For 22 nanometer processors, 1.35 Vcore is above the degradation (electromigration) curve. Even when running big air or 240mm or larger AIOs, at 100% workload overclocked 22 nanometer processors become increasingly difficult to cool as Vcore approaches 1.300. This is because unlike 2nd Generation and earlier processors which have a soldered IHS, 3rd through 8th Generation processors instead have TIM between the IHS and the Die. This is why "delidding" became a solution for overclockers.

Remember to keep overclocking in perspective. For example, the difference between 4.5 GHz and 4.6 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.

Although "Throttle" temperature for the i7-3770K is 105°C, it’s not advisable to run your CPU near its thermal limit. The consensus among well informed and highly experienced system builders, reviewers and overclockers, is that cooler is better for ultimate stability, performance and longevity. Experts all agree that it's prudent to observe a reasonable thermal margin below Throttle temperature, so here's the nominal operating range for Core temperature:

Core temperatures above 85°C are not recommended.

Core temperatures below 80°C are ideal.



Keep in mind that Core temperatures increase and decrease with ambient (room) temperature, for which the international "standard" for "normal" room temperature is 22°C or 72°F.

I suggest that you consider setting your Vcore to the maximum recommended value of 1.300 and configuring your highest stable overclock accordingly.

We have a "Sticky" that covers this topic and is recommended reading: Intel Temperature Guide.

CT :sol:
Hey, thanks for that reply. My voltage was actually 1,33, not 1,35. My pc took a break, and now its running pretty fine. temps at Witcher 3 are under 90 degrees. I am going to change thermal paste asap. But how tight should i mount my AIO(to the cpu i mean). Should it be really tight, or not so tight? Cause many people said that mount fails could cause problems.
 

toniplavna

Reputable
Dec 4, 2016
138
16
4,615
Which "benchmark"?
At what ambient temperature?


Your AIO may be prematurely failing.


Prime95 v29.8 Small FFT's, all AVX selections disabled.

Guys,

Respectfully, you've both apparently been misinformed concerning Core voltage. Processors of different microarchitectures vary in voltage tolerances, which drives thermal behavior, as well as longevity. Regardless of what you may have "read" or "heard", it's likely that you've been misled into thinking that "1.35 is not that much". While this is true for 14 nanometer processors, 1.35 is indeed too much for 22 nanometer processors. Accordingly, a "one size fits all" approach does not apply to maximum recommended Core voltage.

No two processors are identical; each is unique in voltage tolerance, thermal behavior and overclocking potential. Overclocking is always limited by two factors; voltage and temperature.

We know that over time, excessive voltage and temperature can damage electronics. In CPUs, this can result in accelerated "Electromigration" - https://www.google.com/?gws_rd=ssl#q=Electromigration

This prematurely erodes the traces and junctions within the processor's layers and nano-circuits, which will eventually result in blue-screen crashes that become increasingly frequent over time. As a rule, CPU's are more susceptible to Electromigration with each Die-shrink. However, the most notable exception is Intel's 14 nanometer Microarchitecture, where advances in FinFET transistor technology have improved voltage tolerance, which does not include 22 nanometer 3rd Generation processors such as the i7-3770K.

Here's the maximum recommended Core voltage per Microarchitecture from 14 to 65 nanometers since 2006:



For 22 nanometer processors, 1.35 Vcore is above the degradation (electromigration) curve. Even when running big air or 240mm or larger AIOs, at 100% workload overclocked 22 nanometer processors become increasingly difficult to cool as Vcore approaches 1.300. This is because unlike 2nd Generation and earlier processors which have a soldered IHS, 3rd through 8th Generation processors instead have TIM between the IHS and the Die. This is why "delidding" became a solution for overclockers.

Remember to keep overclocking in perspective. For example, the difference between 4.5 GHz and 4.6 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.

Although "Throttle" temperature for the i7-3770K is 105°C, it’s not advisable to run your CPU near its thermal limit. The consensus among well informed and highly experienced system builders, reviewers and overclockers, is that cooler is better for ultimate stability, performance and longevity. Experts all agree that it's prudent to observe a reasonable thermal margin below Throttle temperature, so here's the nominal operating range for Core temperature:

Core temperatures above 85°C are not recommended.

Core temperatures below 80°C are ideal.



Keep in mind that Core temperatures increase and decrease with ambient (room) temperature, for which the international "standard" for "normal" room temperature is 22°C or 72°F.

I suggest that you consider setting your Vcore to the maximum recommended value of 1.300 and configuring your highest stable overclock accordingly.

We have a "Sticky" that covers this topic and is recommended reading: Intel Temperature Guide.

CT :sol:

Agree. He should, after changing paste and checking cooler, lower the OC to maybe 4.4ghz and drop the Vcore below 1.3v
 
  • Like
Reactions: CompuTronix

rodrigoxm49

Great
Oct 13, 2019
119
8
95
Which "benchmark"?
At what ambient temperature?


Your AIO may be prematurely failing.


Prime95 v29.8 Small FFT's, all AVX selections disabled.

Guys,

Respectfully, you've both apparently been misinformed concerning Core voltage. Processors of different microarchitectures vary in voltage tolerances, which drives thermal behavior, as well as longevity. Regardless of what you may have "read" or "heard", it's likely that you've been misled into thinking that "1.35 is not that much". While this is true for 14 nanometer processors, 1.35 is indeed too much for 22 nanometer processors. Accordingly, a "one size fits all" approach does not apply to maximum recommended Core voltage.

No two processors are identical; each is unique in voltage tolerance, thermal behavior and overclocking potential. Overclocking is always limited by two factors; voltage and temperature.

We know that over time, excessive voltage and temperature can damage electronics. In CPUs, this can result in accelerated "Electromigration" - https://www.google.com/?gws_rd=ssl#q=Electromigration

This prematurely erodes the traces and junctions within the processor's layers and nano-circuits, which will eventually result in blue-screen crashes that become increasingly frequent over time. As a rule, CPU's are more susceptible to Electromigration with each Die-shrink. However, the most notable exception is Intel's 14 nanometer Microarchitecture, where advances in FinFET transistor technology have improved voltage tolerance, which does not include 22 nanometer 3rd Generation processors such as the i7-3770K.

Here's the maximum recommended Core voltage per Microarchitecture from 14 to 65 nanometers since 2006:



For 22 nanometer processors, 1.35 Vcore is above the degradation (electromigration) curve. Even when running big air or 240mm or larger AIOs, at 100% workload overclocked 22 nanometer processors become increasingly difficult to cool as Vcore approaches 1.300. This is because unlike 2nd Generation and earlier processors which have a soldered IHS, 3rd through 8th Generation processors instead have TIM between the IHS and the Die. This is why "delidding" became a solution for overclockers.

Remember to keep overclocking in perspective. For example, the difference between 4.5 GHz and 4.6 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.

Although "Throttle" temperature for the i7-3770K is 105°C, it’s not advisable to run your CPU near its thermal limit. The consensus among well informed and highly experienced system builders, reviewers and overclockers, is that cooler is better for ultimate stability, performance and longevity. Experts all agree that it's prudent to observe a reasonable thermal margin below Throttle temperature, so here's the nominal operating range for Core temperature:

Core temperatures above 85°C are not recommended.

Core temperatures below 80°C are ideal.



Keep in mind that Core temperatures increase and decrease with ambient (room) temperature, for which the international "standard" for "normal" room temperature is 22°C or 72°F.

I suggest that you consider setting your Vcore to the maximum recommended value of 1.300 and configuring your highest stable overclock accordingly.

We have a "Sticky" that covers this topic and is recommended reading: Intel Temperature Guide.

CT :sol:

I remember very well that Intel says that 1.38v is the maximum of Vcore for Sandy Bridge, for example. I remember that a very know site find Intel documents that claims 1.425v as the max recommended. But let's stick with oficial 1.38v.

I managed to run for almost a decade a 2500k with 1.34v. 120mm AIO was more than enough to keep my temps in a good term on gaming (60C max). So it's not like "a guy is saying on reddit". It was my experience. For years and years and years working 24/7. But you can search for it and you will find a lot of guys saying the same thing too. Because everyone knows that 1.30v is the basics of OC on these CPUs.

Sadly I could not find the oficial Intel article, but you can find in all internet people using 1.38v or even 1.45 as a parameter and quoting about it.

So, 1.30v is absolutley safe.
 

CompuTronix

Intel Master
Moderator
... Intel says that 1.38v is the maximum of Vcore for Sandy Bridge ... Intel documents that claims 1.425v as the max recommended. But let's stick with oficial 1.38v ... 2500k with 1.34 ... everyone knows that 1.30v is the basics of OC on these CPUs ... I could not find the oficial Intel article ... all internet people using 1.38v or even 1.45 as a parameter and quoting about it.

So, 1.30v is absolutley safe.

rodrigoxm49,

Respectfully, while your last statement is true, let's not confuse apples with oranges, lest we have Vcore fruit salad in a blender. The O.P.'s interest concerns his 3rd Generation 22 nanometer Ivy Bridge i7-3770K.

However, with regard to your 2nd Generation 32 nanometer Sandy Bridge i5-2500K, here's Intel's Datasheet: 2nd Generation Intel® Core™ Processor Family Desktop, Datasheet, Volume 1.

See page 80, Section 7.10.1, Voltage and Current Specifications, Table 7-5, Row 1, Column 5: Maximum 1.52 volts. The value shown in the table I provided in my previous post shows 1.375, which is very close to your claim of 1.38. Here's the table again:



With regard to boriss911's 3rd Generation 22 nanometer Ivy Bridge i7-3770K, the Datasheet (page 84) also shows the same voltage, while for 4th Generation 22 nanometer Haswell (and Devil's Canyon) processors, the Datasheet (page 102) shows Maximum 1.86 volts. The value shown in the table shows 1.30.

Let's keep in mind that while Intel's Datasheets have authority, they nevertheless contain certain inconsistencies, which may or may not be superceded in their Errata documentation, and may or may not be non-proprietary information.

For example, the Datasheets for Core i 1st Generation 45nm processors as well contain Voltage and Current Specifications which shows VID as 1.375, yet a section called Processor Absolute Minimum and Maximum Ratings shows 1.55 volts. This section does not appear in later Datasheets. Intel does not typically show absolute Vcore, but instead shows VID. Most users who glance through the Tables in the Datasheets commonly misinterpret VID for Vcore, which is how misinformation proliferates faster than rabbits.

There are many instances where earlier Datasheets include applicable information that later Datasheets do not, or vice versa, as if at certain junctures Intel thought better of allowing non-proprietary access to certain bits of information. The best means by which anyone can achieve a global understand of Intel's various specifications and how they relate to one another, is to devote the time necessary to sit down and not just glance, skim or even read, but study all their Datasheets, and the differences as well as the inconsistencies among various Generations.

Taken together with respect to microarchitecture, VID values include:

14nm
6th thru 9th Gen - 1.52
* 5th Gen - 1.86

22nm
* 4th Gen - 1.86
3rd Gen - 1.52

32nm
2nd Gen - 1.52

45nm
1st Gen - 1.375
or Max - 1.55

* Denotes processors with Fully Integrated Voltage Regulators (FIVR).

Obviously, Intel's values differ greatly from the table shown above. As per the collective knowledge of well informed and highly experienced system builders, reviewers and overclockers, all of these voltage values, except for one, will not only be impossible to cool, but they will quickly result in degradation; especially a glaring value like 1.86 volts on a 4th Gen Haswell.

I suggest that you watch this YouTube which is an in-depth explanation of VID and Vcore:


We also know that that exceedingly high Core voltages offer no practical advantage to overclocking, as most processors, in addition to having a thermal "sweet spot" also have a frequency "wall", where regardless of cooling, no amount of additional Vcore will allow a higher overclock, and instead will often result in destabilizing an overclock.

While the Maximum Recommended Vcore values shown in the table above may seem to be a bit on the conservative side to a few users, there remains a very important additional consideration ... out of respect for the hard-earned personal property of others, we can not, in good conscience, recommend higher Core voltages that extend beyond the threshold values where accelerated processor or VRM degradations are known to occur.

CT :sol:
 

CompuTronix

Intel Master
Moderator
boriss911 said:
Well, its still pretty hot. Can lo voltage damage CPU?
Low Core voltage will not damage your processor, however, it can typically results in repeated Blue Screen crashes (BSOD), which can cause a corrupted Operating System (OS) and / or corrupted applications software, not to mention losing precious data files. As such, it's always strongly recommended to regularly maintain multiple full system backups, especially when testing overclock stability. Moreover, repeated crashes due to an unstable memory overclock makes software particularly vulnerable to being corrupted, which amounts to binary Russian Roulette.

To reiterate, the best and most responsible overclocking advice I can give anyone, is to keep it in perspective. For example, the difference between 4.5 GHz and 4.6 Ghz is less than 2.3%, which has no noticeable impact on overall system performance. If common sense prevails, then it simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits, and sacrificing longevity for degradation, just to squeeze out another 100 MHz.

CT :sol:
 
Do it. If nothing changes, maybe you need to delid it.

Delidding is never the ANSWER to ANY thermal problems. Delidding should ONLY be used when an overclocker wants to get MORE out of an already 100% compliant CPU that is hitting a thermal wall and cannot be overclocked further because of it. We don't delid to fix thermal issues with stock configurations. Please, don't offer that as a troubleshooting step in the future because it just is bad advice and plain wrong.
 
I can tell you, for a fact, that practically every one of those ML240L cooler threads I've seen here and on other forums that has had thermal issues was due to a faulty pump. The L in 240L stands for "lite". Any cooler or power supply with "lite" in it's name, should be avoided as a matter of course. Given your random issues I'd say you either have a faulty pump or an air bubble trapped in the pump, if you are still having thermal issues.

As Computronix has already stated, running lower voltage is not the solution. All that will do is create an unstable configuration. You need to run the CORRECT voltage for whatever clock speed you are at. WHAT that voltage is, can be determined two ways. One way is to reset the BIOS to it's default settings and run it at the stock configuration. The other way is through trial and error, with testing for stability, but that cannot be done if it will not even maintain thermal compliance at the stock configuration.

No system should ever have a problem running at the stock configuration IF you are using a motherboard that is adequate for the CPU being used, AND if you are using a cooler that is at least as good as the stock cooler it came with, if not better. For CPUs that did not come with a stock cooler, then a cooler that is rated for at least the TDP of the processor. In your case, it should be more than sufficient to keep it from having any thermal issues at the stock/default configuration, IF it is working properly. It may NOT be working properly.

For my part, I dislike the use of Speedfan. I know some people still hold on to the use of it because it does have some capabilities not found in most older BIOS or desktop applications, but I've seen it cause just as many problems if not more, than it cures. And, more often than not settings changed using speedfan tend to not want to reverse or be undone when you try to change them back so a full reset of the BIOS and complete reinstallation of Windows is sometimes necessary in order to do so.

In any case, I don't believe Speedfan should be used for control of any portion of an AIO cooler. I'm sure not everybody agrees with that, and that's fine. Personally I think you're problem is likely related to either a pump that is faulty or a pump or fans that are not running at the proper speed.

I'll tell you for sure, that cooler is one of the worst 240mm AIO coolers you could get and I'd replace it if you have any thoughts of managing to get your overclock thermally compliant. Even if nothing is wrong with it at all.

I would start by resetting EVERYTHING to the stock settings, and see where you are at in terms of thermals then. If there is a problem with temperatures when everything is at the default configuration, then there is a problem, period, and it needs to get addressed. If there is not a problem once you return everything to the stock settings, THEN you can start figuring out your overclock little by little. You can't just jump to 4.6Ghz at X voltage and think it's going to work out. You need to raise the multi and test. If it's unstable, you increase voltage and restest until it is stable again AND within thermal limits. Rinse and repeat.

But FIRST you need to know if there is a cooling problem at the baseline, otherwise you are just wasting your time, and everybody else's.
 
Last edited:

boriss911

Commendable
Nov 1, 2019
68
8
1,535
I can tell you, for a fact, that practically every one of those ML240L cooler threads I've seen here and on other forums that has had thermal issues was due to a faulty pump. The L in 240L stands for "lite". Any cooler or power supply with "lite" in it's name, should be avoided as a matter of course. Given your random issues I'd say you either have a faulty pump or an air bubble trapped in the pump, if you are still having thermal issues.

As Computronix has already stated, running lower voltage is not the solution. All that will do is create an unstable configuration. You need to run the CORRECT voltage for whatever clock speed you are at. WHAT that voltage is, can be determined two ways. One way is to reset the BIOS to it's default settings and run it at the stock configuration. The other way is through trial and error, with testing for stability, but that cannot be done if it will not even maintain thermal compliance at the stock configuration.

No system should ever have a problem running at the stock configuration IF you are using a motherboard that is adequate for the CPU being used, AND if you are using a cooler that is at least as good as the stock cooler it came with, if not better. For CPUs that did not come with a stock cooler, then a cooler that is rated for at least the TDP of the processor. In your case, it should be more than sufficient to keep it from having any thermal issues at the stock/default configuration, IF it is working properly. It may NOT be working properly.

For my part, I dislike the use of Speedfan. I know some people still hold on to the use of it because it does have some capabilities not found in most older BIOS or desktop applications, but I've seen it cause just as many problems if not more, than it cures. And, more often than not settings changed using speedfan tend to not want to reverse or be undone when you try to change them back so a full reset of the BIOS and complete reinstallation of Windows is sometimes necessary in order to do so.

In any case, I don't believe Speedfan should be used for control of any portion of an AIO cooler. I'm sure not everybody agrees with that, and that's fine. Personally I think you're problem is likely related to either a pump that is faulty or a pump or fans that are not running at the proper speed.

I'll tell you for sure, that cooler is one of the worst 240mm AIO coolers you could get and I'd replace it if you have any thoughts of managing to get your overclock thermally compliant. Even if nothing is wrong with it at all.

I would start by resetting EVERYTHING to the stock settings, and see where you are at in terms of thermals then. If there is a problem with temperatures when everything is at the default configuration, then there is a problem, period, and it needs to get addressed. If there is not a problem once you return everything to the stock settings, THEN you can start figuring out your overclock little by little. You can't just jump to 4.6Ghz at X voltage and think it's going to work out. You need to raise the multi and test. If it's unstable, you increase voltage and restest until it is stable again AND within thermal limits. Rinse and repeat.

But FIRST you need to know if there is a cooling problem at the baseline, otherwise you are just wasting your time, and everybody else's.
pump is running constantly at 2.3k rpm. I feel one tube vibrating when i touch it. I think problem is just in 3770k, as they are known for very high temps.
 

CompuTronix

Intel Master
Moderator
... using ml240lrgb ... about a year ... 35 idle, 80 gaming ... cpu went up to 100 ... cpu temps wont go down ... going to remount aio, and change thermal paste.

boriss911,

Thermal compound fails slowly over a long period of time (3 years or more) and will not cause Core temperatures to suddenly increase. You apparently have an intermittent flow problem, which is typical for many AIO's, even those which have only been in service for a year.

I moved my pc abit, so maybe AIO got loose.

pump is running constantly at 2.3k rpm. I feel one tube vibrating when i touch it. I think problem is just in 3770k, as they are known for very high temps.

It's more likely that by simply moving your case, it's caused an air bubble to migrate from the radiator into the impeller chamber, just as Darkbreeze has already suggested. Read on ...

Apart from the potential for a coolant leak, there are several problems common to AIO's:

All AIO's will eventually fail. It’s not a question of if; it’s a question of when. Pumps can have component failures in the electronics that drive the impeller, which is a moving part that can wear out, so those which run 24/7/365 are prone to premature failure. AIO units are notorious for failures due to inferior pump quality, whereas custom loops typically use high-end pumps which have greater longevity.

Coolant can slowly evaporate over time due to "permeation". However unlikely it may seem, coolant can actually dissipate directly through the tubes. This can introduce bubbles into the impeller chamber causing "cavitation", whereby coolant flow is impeded or interrupted.

AIO's are sealed Closed Loop Coolers (CLC), which unlike custom loops, are not designed to be disassembled for the water block and impeller to be flushed, cleaned of bio-contaminants and radiator sediments, then refilled. Bio-contaminants gradually accumulate over time and will eventually clog the water block and radiator tubes, thereby reducing flow and thermal efficiency. This is evidenced by Core temperatures that slowly increase over periods of several months to a few years.

Also unlike custom loops, AIO's use dissimilar metals (aluminum radiator / copper water block). This causes galvanic corrosion which produces sediments that accumulate over time, resulting in blockages and flow problems. Even new AIO's may contain radiator sediments due to inadequate flushing after manufacturing. Experienced builders of custom loops will always thoroughly flush brand-new radiators. Flux, solder and metal fragments are typically found in the flush water when it's poured through a strainer.

For no apparent reason ... OR ... by simply installing, moving, tipping, handling or otherwise disturbing a NEW OR OLD unit, particles can become dislodged, whereupon the next power-up, the now free-floating particles can circulate into the impeller chamber and restrict or completely stop the impeller. Since the impeller is magnetically coupled to the stator (no direct shaft), the unit may "appear" that it's running while the impeller is restricted or stalled. "Hearing" the unit running or "feeling" vibration can be deceiving as it does not necessarily indicate flow, nor does Pump RPM in BIOS or various software utilities. Moreover, as fan vibration can "telegraph" throughout the entire AIO unit, it's often mistaken for pump vibration.

Without an actual in-line sensor, proper flow is difficult to confirm. Under normal operation, even with the CPU at 100% workload, there should be only a minimal temperature differential between the tubes. However, if one tube is hot while the other is cool, or the water block is hot while the radiator is cool, it indicates little to no flow. Surface temperatures can be verified with an infrared (IR) thermometer.

Since your AIO is apparently only about a year old, I suggest that you submit an RMA request for a warranty replacement. In the interim, the original stock air cooler should suffice so your rig won't be down.

CT :sol:
 

boriss911

Commendable
Nov 1, 2019
68
8
1,535
boriss911,

Thermal compound fails slowly over a long period of time (3 years or more) and will not cause Core temperatures to suddenly increase. You apparently have an intermittent flow problem, which is typical for many AIO's, even those which have only been in service for a year.





It's more likely that by simply moving your case, it's caused an air bubble to migrate from the radiator into the impeller chamber, just as Darkbreeze has already suggested. Read on ...

Apart from the potential for a coolant leak, there are several problems common to AIO's:

All AIO's will eventually fail. It’s not a question of if; it’s a question of when. Pumps can have component failures in the electronics that drive the impeller, which is a moving part that can wear out, so those which run 24/7/365 are prone to premature failure. AIO units are notorious for failures due to inferior pump quality, whereas custom loops typically use high-end pumps which have greater longevity.

Coolant can slowly evaporate over time due to "permeation". However unlikely it may seem, coolant can actually dissipate directly through the tubes. This can introduce bubbles into the impeller chamber causing "cavitation", whereby coolant flow is impeded or interrupted.

AIO's are sealed Closed Loop Coolers (CLC), which unlike custom loops, are not designed to be disassembled for the water block and impeller to be flushed, cleaned of bio-contaminants and radiator sediments, then refilled. Bio-contaminants gradually accumulate over time and will eventually clog the water block and radiator tubes, thereby reducing flow and thermal efficiency. This is evidenced by Core temperatures that slowly increase over periods of several months to a few years.

Also unlike custom loops, AIO's use dissimilar metals (aluminum radiator / copper water block). This causes galvanic corrosion which produces sediments that accumulate over time, resulting in blockages and flow problems. Even new AIO's may contain radiator sediments due to inadequate flushing after manufacturing. Experienced builders of custom loops will always thoroughly flush brand-new radiators. Flux, solder and metal fragments are typically found in the flush water when it's poured through a strainer.

For no apparent reason ... OR ... by simply installing, moving, tipping, handling or otherwise disturbing a NEW OR OLD unit, particles can become dislodged, whereupon the next power-up, the now free-floating particles can circulate into the impeller chamber and restrict or completely stop the impeller. Since the impeller is magnetically coupled to the stator (no direct shaft), the unit may "appear" that it's running while the impeller is restricted or stalled. "Hearing" the unit running or "feeling" vibration can be deceiving as it does not necessarily indicate flow, nor does Pump RPM in BIOS or various software utilities. Moreover, as fan vibration can "telegraph" throughout the entire AIO unit, it's often mistaken for pump vibration.

Without an actual in-line sensor, proper flow is difficult to confirm. Under normal operation, even with the CPU at 100% workload, there should be only a minimal temperature differential between the tubes. However, if one tube is hot while the other is cool, or the water block is hot while the radiator is cool, it indicates little to no flow. Surface temperatures can be verified with an infrared (IR) thermometer.

Since your AIO is apparently only about a year old, I suggest that you submit an RMA request for a warranty replacement. In the interim, the original stock air cooler should suffice so your rig won't be down.

CT :sol:
Water block is cool, one side of radiator is warmer that the other and one tube vibrates a bit more than the other. How do i actually get rid of air bubble?
 

rodrigoxm49

Great
Oct 13, 2019
119
8
95
rodrigoxm49,

Respectfully, while your last statement is true, let's not confuse apples with oranges, lest we have Vcore fruit salad in a blender. The O.P.'s interest concerns his 3rd Generation 22 nanometer Ivy Bridge i7-3770K.

However, with regard to your 2nd Generation 32 nanometer Sandy Bridge i5-2500K, here's Intel's Datasheet: 2nd Generation Intel® Core™ Processor Family Desktop, Datasheet, Volume 1.

See page 80, Section 7.10.1, Voltage and Current Specifications, Table 7-5, Row 1, Column 5: Maximum 1.52 volts. The value shown in the table I provided in my previous post shows 1.375, which is very close to your claim of 1.38. Here's the table again:



With regard to boriss911's 3rd Generation 22 nanometer Ivy Bridge i7-3770K, the Datasheet (page 84) also shows the same voltage, while for 4th Generation 22 nanometer Haswell (and Devil's Canyon) processors, the Datasheet (page 102) shows Maximum 1.86 volts. The value shown in the table shows 1.30.

Let's keep in mind that while Intel's Datasheets have authority, they nevertheless contain certain inconsistencies, which may or may not be superceded in their Errata documentation, and may or may not be non-proprietary information.

For example, the Datasheets for Core i 1st Generation 45nm processors as well contain Voltage and Current Specifications which shows VID as 1.375, yet a section called Processor Absolute Minimum and Maximum Ratings shows 1.55 volts. This section does not appear in later Datasheets. Intel does not typically show absolute Vcore, but instead shows VID. Most users who glance through the Tables in the Datasheets commonly misinterpret VID for Vcore, which is how misinformation proliferates faster than rabbits.

There are many instances where earlier Datasheets include applicable information that later Datasheets do not, or vice versa, as if at certain junctures Intel thought better of allowing non-proprietary access to certain bits of information. The best means by which anyone can achieve a global understand of Intel's various specifications and how they relate to one another, is to devote the time necessary to sit down and not just glance, skim or even read, but study all their Datasheets, and the differences as well as the inconsistencies among various Generations.

Taken together with respect to microarchitecture, VID values include:

14nm
6th thru 9th Gen - 1.52
* 5th Gen - 1.86

22nm
* 4th Gen - 1.86
3rd Gen - 1.52

32nm
2nd Gen - 1.52

45nm
1st Gen - 1.375
or Max - 1.55

* Denotes processors with Fully Integrated Voltage Regulators (FIVR).

Obviously, Intel's values differ greatly from the table shown above. As per the collective knowledge of well informed and highly experienced system builders, reviewers and overclockers, all of these voltage values, except for one, will not only be impossible to cool, but they will quickly result in degradation; especially a glaring value like 1.86 volts on a 4th Gen Haswell.

I suggest that you watch this YouTube which is an in-depth explanation of VID and Vcore:


We also know that that exceedingly high Core voltages offer no practical advantage to overclocking, as most processors, in addition to having a thermal "sweet spot" also have a frequency "wall", where regardless of cooling, no amount of additional Vcore will allow a higher overclock, and instead will often result in destabilizing an overclock.

While the Maximum Recommended Vcore values shown in the table above may seem to be a bit on the conservative side to a few users, there remains a very important additional consideration ... out of respect for the hard-earned personal property of others, we can not, in good conscience, recommend higher Core voltages that extend beyond the threshold values where accelerated processor or VRM degradations are known to occur.

CT :sol:
I know this is all technical stuff, probably with a great safe margin from Intel or whatever who have made it, but the real life is completely different. If you look on overclockersclub or others overclockers forums, you will see people running 1.36v on 3770k or 3570k since the launch. No problems at all. No damage. Still the same performance as ever.

Anyway, I keep my knowledge by talking with people that have OC day by day for yeas. 1.30v is not only safe, but very far from the safe Vcore for any IB or SB. That said, both are basically the same CPU with very minor and insignificant difference.

Thanks for all this information, anyway! But some times real life is just very very very different from theory. Specially when we're talking about companies worried about their products and warranty.
 

rodrigoxm49

Great
Oct 13, 2019
119
8
95
Delidding is never the ANSWER to ANY thermal problems. Delidding should ONLY be used when an overclocker wants to get MORE out of an already 100% compliant CPU that is hitting a thermal wall and cannot be overclocked further because of it. We don't delid to fix thermal issues with stock configurations. Please, don't offer that as a troubleshooting step in the future because it just is bad advice and plain wrong.
First thing, please stop saying people what they should do and what they not should do here. Just give your opinion about whatever you think you should.

About what you say, free lesson from a guy that works with it: delid is not recommended only for people that need extreme OC, but for people that are having huge temps issues with used CPU and no cooler solution was able to fix it. This is very very very common on old i5 and old i7. Many people used for years theses CPUs with Intel coolerbox, reaching 90C/95C on gaming for years and years. Over time it can destroy the thermal compound inside CPU. Really common.

I dont have recommended for OP to delid. I say what everyones that really know and work with this would said: if nothings fix it, maybe,and just maybe, the thermal compound inside is dead.
 

boriss911

Commendable
Nov 1, 2019
68
8
1,535
First thing, please stop saying people what they should do and what they not should do here. Just give your opinion about whatever you think you should.

About what you say, free lesson from a guy that works with it: delid is not recommended only for people that need extreme OC, but for people that are having huge temps issues with used CPU and no cooler solution was able to fix it. This is very very very common on old i5 and old i7. Many people used for years theses CPUs with Intel coolerbox, reaching 90C/95C on gaming for years and years. Over time it can destroy the thermal compound inside CPU. Really common.

I dont have recommended for OP to delid. I say what everyones that really know and work with this would said: if nothings fix it, maybe,and just maybe, the thermal compound inside is dead.
But is 90 ish actually dangerous? Cause its like 40 degrees on idle. With open and cold room it was actually running at 25 on idle.
 
First thing, please stop saying people what they should do and what they not should do here. Just give your opinion about whatever you think you should.

About what you say, free lesson from a guy that works with it: delid is not recommended only for people that need extreme OC, but for people that are having huge temps issues with used CPU and no cooler solution was able to fix it. This is very very very common on old i5 and old i7. Many people used for years theses CPUs with Intel coolerbox, reaching 90C/95C on gaming for years and years. Over time it can destroy the thermal compound inside CPU. Really common.

I dont have recommended for OP to delid. I say what everyones that really know and work with this would said: if nothings fix it, maybe,and just maybe, the thermal compound inside is dead.

Everything you've said, so far, in this thread, is 100% wrong. Just because people, even a LOT of people, "do" something, does not mean that what they are doing is "right" or "recommended". I assure you that myself and Computronix have done more overclocking on more architectures for more years than 90% of those people you are reading about and in all probability, than you as well. FWIW, Computronix is also the AUTHOR of the Intel temperature guide, has spent several thousand hours working with, testing and compiling information about all of the various Intel Core-i architectures, as seen here:



If you have some results from some extensive testing you've done, in this kind of detail, that is indisputable emperical evidence which contradicts what is in the Intel data sheets and Comp's Intel temperature guide, I'm sure we would all be glad to see it. Until then, it's all just white noise.

@boriss911 , yes, 90°C is undesirable. Anything over 85°C, brings you to or at least CLOSER to, the point where you start experiencing electromigration/thermal degradation. Intel CPUs "can" run up to 100°C without any appreciable reduction in performance, but that does not mean that by DOING so, you are not also dramatically reducing the lifespan of the CPU. Electromigration is real, and there is ONLY ONE way to reduce it, and that is by keeping operations within the recommended "safe zone" which for Intel Core-I processors happens to be a maximum of 85°C with 80°C being MUCH preferred for long term use.

This is not an opinion. This is not conjecture. This is not something we simply read someplace on Reddit and decided to believe in. This is a PURE FACT. Nobody can make somebody else believe something that they haven't seen with their own eyes, if they are unwilling, but it does not CHANGE the fact.