Overclocking Temperature

[AdmiralFace]

I know that overclocking can dramatically increase operating temperature, but I'd like to know roughly by how much? And are there any other detrimental side effects of over-overclocking?

 

[fffblackmage]

In general, simply increasing clock speed won't increase temperature much. On the other hand, increasing voltage will dramatically raise temperature. How much depends on the combination of clockspeed, voltage, and what cooler you're using. Obviously, a crappy cooler means higher temps, which is why an aftermarket cooler is typically recommended for overclocking.

Side effects... if you don't increase voltage, you may experience instability (BSODs) but usually won't damage the cpu. Increasing voltage is much risker, since you can potentially fry the CPU. The voltage increase is often needed in order to increase clockspeed further.

 

[AdmiralFace]

Thanks, so overclocking my 2.8 GHz cpu to... say 3.6 GHz and increasing the voltage to whatever necessary to do so with custom liquid cooling would be fine then?

 

[Arsynic]

Liquid cooling isn't necessary. A top-notch air cooler may be just what you need.

 

[cubeless]

unless the custom liquid cooling is a patio mister spraying on the stock hsf, that's what such a rig is for...

always use temp shutdown in bios to be sure you don't cook the chip... and be ready to hit the psu off switch (case power off may not work) when testing the first time or two... better safe than smoked...

 

[AdmiralFace]

Liquid cooling isn't necessary

It will be.

increasing voltage increases heat given off, potentially frying the cpu. so with good enough cooling, could I overclock that 2.8 to.. say 5? or does the voltage increase otherwise effect the processor?

lways use temp shutdown in bios to be sure you don't cook the chip...

I have overheating covered, or will do when the PCs built

 

[cubeless]

you need phase change to get to 5...

what chip are we talking about? an i7 may make 3.6 with very little extra voltage, and a good aircooler would suffice...

i have a 930 @ 4.0 @ 1.35v that is ok with a core contact freezer with 2 fans on it... to go to 4.2 it takes 1.4 v, and overheats with that hsf...

i just bought a noctua 14cm crazycooler to see how far air cooling can go, since i hate the hassle of water cooling...

 

[AdmiralFace]

what chip are we talking about?

yeah... thats the problem, atm i have no clue what processor it is, but Iknow its old. its an intel celeron something.

I'm gonna see how cheaply I can make a halfway decent computer, so a cheapy mobo is coming from ebay with a celeron processor. I'm gonna overclock it as much as I can, ignoring heating issues to build custom "liquid cooling".

Little off topic there...

 

[Voo]

If you're interested in the formulas:

static power consumption: transistors are not ideal and do not switch off completly, resulting in a small leakage current that flows even in the off state
dynamic power consumption: gates and wires have capacities that must be charged or discharged during the switching process

f.. frequency
C.. wire capacity

Pdyn: 1/2 * C * Vdd^2 * f
Psta: Idd * Vdd

P = Pdyn + Psta

 

[cubeless]

yeah... thats the problem, atm i have no clue what processor it is, but Iknow its old. its an intel celeron something.

I'm gonna see how cheaply I can make a halfway decent computer, so a cheapy mobo is coming from ebay with a celeron processor. I'm gonna overclock it as much as I can, ignoring heating issues to build custom "liquid cooling".

Little off topic there...

umm, m'kay... good news is that a 'cheapy mobo' will probably save you the trouble of worrying about cooling - you won't be able to oc enuf for it to matter... have fun...

 

[AdmiralFace]

umm, m'kay... good news is that a 'cheapy mobo' will probably save you the trouble of worrying about cooling - you won't be able to oc enuf for it to matter... have fun...

That's what you think.

If its not up to scratch, then I'll go further into wallet.

 

[Ben90]

Thanks, so overclocking my 2.8 GHz cpu to... say 3.6 GHz and increasing the voltage to whatever necessary to do so with custom liquid cooling would be fine then?

NO! I'm surprised no one has said anything out of all these posts.

Raising voltage too far will have devastating effects on a processor. While it can be okay to raise 10-15%, much more than that will ruin a processor regardless of temperature.

I would research what a safe maximum voltage is for your processor before attempting a massive overclock.

 

[AdmiralFace]

Cheers ben90. i was trying to get at that earlier.

 

[Markfw]

every chip series has a maximum. We need to know the chip before we can tell you safe vcore.

 

[mechwarrior1989]

That's what you think.

If its not up to scratch, then I'll go further into wallet.

What exactly are you trying to accomplish? So you'll spend some money for a cheap motherboard and cheap CPU and then when it doesn't overclock properly you'll spend more money on another motherboard? You should see overclocking as a bonus, not as a goal if you're going for a low end CPU. Keep your expectations low.

Many low end motherboards suck for overclocking as people have already mentioned. It all depends on chipset and also how power is distributed. Some motherboard are much better at keeping a constant voltage going into a CPU.

 

[AdmiralFace]

What exactly are you trying to accomplish? So you'll spend some money for a cheap motherboard and cheap CPU and then when it doesn't overclock properly you'll spend more money on another motherboard?

No. Well sort of. I want to buy a cheapish CPu and overclock it, if, after overclocking it still doesn't have the processing power I need, il get a better CPU.

 

[Tlkki]

May i quote mechwarrior:
What exactly are you trying to accomplish?

If you know your processing power needs, there are many a person willing to help you right here to get there as cheap as possible. On the other hand, if you just want to try overclocking with cheap equipment, no one can stop you, but that just isnt very rewarding. Other than being a little bonus as someone mentioned earlier. And cheap ofcourse.

 

[Arsynic]

It will be.

increasing voltage increases heat given off, potentially frying the cpu. so with good enough cooling, could I overclock that 2.8 to.. say 5? or does the voltage increase otherwise effect the processor?



I have overheating covered, or will do when the PCs built

You don't need water for a 3.5 overclock. You could probably get by with an H50 or H70.

 

[AdmiralFace]

You don't need water for a 3.5 overclock. You could probably get by with an H50 or H70.

It's not water, and its a 5.0 overclock, not 3.5.

What exactly are you trying to accomplish?

I wanna see how cheaply I can make a decentish computer: http://mobpcp.blogspot.com/

 

[muskie32]

IF you are the one that posted that... You are going to put the parts in vegetable oil?

 

[AdmiralFace]

IF you are the one that posted that... You are going to put the parts in vegetable oil?

or mineral oil. preferably mineral oil as it doesn't go rancid.

 

[Ben90]

While I hate to be the bearer of bad news, don't expect thriller temps out of a mineral oil computer. They can be an interesting experiment, but can have problems competing with air coolers until you start throwing money into them. From there on, a budget watercooling setup will start to pull out ahead.

There are a few problems with mineral oil computers. The main one is that you have to cool absolutely every component whether you want to or not. VRMs, capacitors, RAM, and even the fans themselves will put out heat that you need to find a way of dispersing. If you plan on overclocking, a passive system will absolutely be out of the question.

I don't want to discourage you from making your own though however as it can be a fun experiment. There is also a huge advantage to a mineral oil computer as well: silence.

The guys on SilentPC.com could spend thousands and thousands of dollars trying to insulate capacitor whine, but the first time you turn a mineral pc on, you get blown away by the absolute 100% lack of noise.

R.I.P. AquaMan

Yesterday a great tragedy befell us. AquaMan vent server has passed away. He lived a hot and oily life, but ceased functioning when a capacitor came loose from his motherboard.​

​

 

[muskie32]

that sounds pretty interesting though. ^_^

 

[cbn]

http://www.youtube.com/watch?v=PtufuXLvOok&feature=player_embedded#!

After you mentioned mineral oil cooling I found this video from Puget Custom computers.

Mineral oil and fans? Why are they being used together? Wouldn't a fan less PSU be more appropriate to this kind of application?

 

[Devilpapaya]

http://www.youtube.com/watch?v=PtufuXLvOok&feature=player_embedded#!

After you mentioned mineral oil cooling I found this video from Puget Custom computers.

Mineral oil and fans? Why are they being used together? Wouldn't a fan less PSU be more appropriate to this kind of application?

If you read Puget's report on their mineral oil project they make note of that. The fans all burnt out after awhile due to the resistance (mechanical) from the mineral oil; so it didn't make much of a difference in the long run.