Do you actually check how the TIM is applied?

[hivebattleii]

Do you actually check how TIM is applied?

Do you guys check whether TIM is properly spread out or not?
Or do you guys never look back once the TIM is applied?

Is it even normal to take away heatsink and check back whether TIM is properly applied?

Sorry. Im new to OC and this is my first time applying thermal paste to cpu.

 

[PVTCaboose1337]

I never look... I just kinda know if i did good or not.

 

[JC316]

I never look. I have gotten so good at it, I just know. Usually, you need a pretty thin layer. Not quite transparent, but close.

 

[miloshs]

I never look also, it's kinda pain in the ass
But u can be sure u put enough if you just put a little dab(or two dabs diagonally across the IHS) of TIM aprox. 1-1,5mm in diameter... Should be enough.

And if your guts is telling you that there's too much... well... your guts is right

 

[JC316]

I never look also, it's kinda pain in the ass
But u can be sure u put enough if you just put a little dab(or two dabs diagonally across the IHS) of TIM aprox. 1-1,5mm in diameter... Should be enough.

And if your guts is telling you that there's too much... well... your guts is right

See, I like to spread mine out before putting the heatsink on. I always wind up with a blob on one side and none on the other when I do that, probably from the push pin system that Intel uses. Since you can't screw it down evenly, it all gets squeezed to one side.

 

[Duxx]

If you take it off and look, u wind up messing it all up and putting too much in certain places, if however you seat it and notice significant temperatures, then id say go ahead and look to see how its layed out.

 

[SK-1]

I never look. I have gotten so good at it, I just know. Usually, you need a pretty thin layer. Not quite transparent, but close.

So many people use way too much. I do the almost transparent thing myself.

 

[AddSub]

No checking here. I have done the procedure so many times (in the triple digits by now) on CPUs, GPUs, and chipset IC's that I pretty much know how much of it needs to be put and how it needs to be done.

 

[hat]

Uhh... I think I did once, the first time I changed nvidia's thermal gunk with some silver stuff. I saw a nice translucent layer on both the gpu core and the heatsink so I was happy.

 

[nanohead]

No, it simply does not matter. There is so much misinformation on the internet, it is silly. People make it out like its some kind of secret sauce that can make computers faster and more mysterious.

The fact is, all of the different types behave much the same, and most of the 'tests' of this stuff are dopey and inconsistent. Its clear there are some that are good, and others that are better, but people get hung on on .01 of a degree (1/100 of a degree!!!)... like the motherboard sensors are even close to that accurate.

You could fart on your motherboard and raise the temp by more than that.

This whole "TIM" thing is more internet insanity, more voodoo computing. Fact is, most production solid state products that need heat sinks use it, and they ALL work fine. My favorite is how these adolescent knuckleheads get a new mobo or vid card, and they "immediately apply arctic silver 5" and then claim how much better everything runs.....

The lapping thing is another idiotic internet frenzy.... More dumbness. The extra degree or so makes ABSOLUTELY NO DIFFERENCE in the real world.

 

[Evo85]

I disagree Nano. I have 2-3 degrees Celsius change from a stock cooler/paste on my 3850 to my Zalman W/AS 5.

When your pushing electronics, heat is the #1 enemy!

As for checking my application job. I did the first few times I used TIP. I learned quickly what was right and wrong.

 

[sneekypeet]

No, it simply does not matter. There is so much misinformation on the internet, it is silly. People make it out like its some kind of secret sauce that can make computers faster and more mysterious.

The fact is, all of the different types behave much the same, and most of the 'tests' of this stuff are dopey and inconsistent. Its clear there are some that are good, and others that are better, but people get hung on on .01 of a degree (1/100 of a degree!!!)... like the motherboard sensors are even close to that accurate.

You could fart on your motherboard and raise the temp by more than that.

This whole "TIM" thing is more internet insanity, more voodoo computing. Fact is, most production solid state products that need heat sinks use it, and they ALL work fine. My favorite is how these adolescent knuckleheads get a new mobo or vid card, and they "immediately apply arctic silver 5" and then claim how much better everything runs.....

The lapping thing is another idiotic internet frenzy.... More dumbness. The extra degree or so makes ABSOLUTELY NO DIFFERENCE in the real world.

This comment is very narrowminded and way inside the box. If you were to be they type to OC for every little bit, Lapping and correct TIM application can mean as much as 10*c drop after. From someone who has done the blob and the spread method, as well as lapping a few things in my PC lifetime, I can say it ends up being very worth it.

For those of us who can afford phasechange, water is the next best thing. If left on air cooling alone, I lapped everything.

Also lapping a DFI NF4 chipset cooler and applying a quality TIM landed me 15*c cooler at idle and 10*c cooler at load.

I quess Im wrong tho, it must have been the temperature fairies blessing me that day!

 

[Squirrely]

I have to say, the temperature fairies do help out every once in a while! Especially when you've done so well with the cooling that a C2D for example reports under ambient temps.

But, thermal paste can make all the difference. For example, in my laptop (has a T5600 and a 7400 Go) was 15C cooler on the processor and 10C on the graphics card on average under load and idle, by just having MX-2 on the heatsink! (There was some crap paste on that one, lol, felt like dried gum) Even on my desktop, MX-2 compared to AS5 is about 2-3C cooler on my GPU and CPU. That's quite a big difference. So, certain techniques such as lapping, and the type of thermal paste add up in big temp differences.

But, back on the topic of the thread, I have done it to so many times, I usually just use the "BB sized dot" method, which spreads very evenly, and I never check. (I did check the first few times, though lol) I have not done a quad yet though. If I was, I would use the thin X method, which seems to work well for quads.

But other than that, temperature fairies ftw!

 

[EastCoasthandle]

I use the line method when applying thermal compound. There is also the pea/dollop, rice method but that doesn't work for me.

There is a review about it called Spreading it Thin TIM Roundup 2007

 

[kenkickr]

I guess I'm a weird one cause I like to take a heatsink off and check out how the TIM is spreading, only when I purchase a 3rd party heatsink cause not all spread alike it seems anymore. With retails or OEM(The ones that come in HP, Dell, etc) I just spread it across the cpu and throw it own.

 

[Pinchy]

The trick is to listen to what the manufacturer says. You may think they just put in that step for fun...but generally what they say works.

For eg, the new DHT coolers (well most of them) require you to put double the amount of TIM as normal...due to the design. You may think its stupid and that excess TIM makes it warmer...but I have tested both ways and the excess has made a diff (not much, but slightly cooler).

No, it simply does not matter. There is so much misinformation on the internet, it is silly. People make it out like its some kind of secret sauce that can make computers faster and more mysterious.

The fact is, all of the different types behave much the same, and most of the 'tests' of this stuff are dopey and inconsistent. Its clear there are some that are good, and others that are better, but people get hung on on .01 of a degree (1/100 of a degree!!!)... like the motherboard sensors are even close to that accurate.

You could fart on your motherboard and raise the temp by more than that.

This whole "TIM" thing is more internet insanity, more voodoo computing. Fact is, most production solid state products that need heat sinks use it, and they ALL work fine. My favorite is how these adolescent knuckleheads get a new mobo or vid card, and they "immediately apply arctic silver 5" and then claim how much better everything runs.....

Thats only a half truth. Some pastes do work better than others by approx 2-3*C (never seen anything more than that).

 

[mrw1986]

No, it simply does not matter. There is so much misinformation on the internet, it is silly. People make it out like its some kind of secret sauce that can make computers faster and more mysterious.

The fact is, all of the different types behave much the same, and most of the 'tests' of this stuff are dopey and inconsistent. Its clear there are some that are good, and others that are better, but people get hung on on .01 of a degree (1/100 of a degree!!!)... like the motherboard sensors are even close to that accurate.

You could fart on your motherboard and raise the temp by more than that.

This whole "TIM" thing is more internet insanity, more voodoo computing. Fact is, most production solid state products that need heat sinks use it, and they ALL work fine. My favorite is how these adolescent knuckleheads get a new mobo or vid card, and they "immediately apply arctic silver 5" and then claim how much better everything runs.....

The lapping thing is another idiotic internet frenzy.... More dumbness. The extra degree or so makes ABSOLUTELY NO DIFFERENCE in the real world.

You couldn't be more wrong. I replaced the TIM on the vregs on my mobo along with my chipset with Ceramique and noticed almost a 10C degree drop. I also did this on my 8800GT's when I had them and it dropped almost 10C. Please don't go spreading crap around which you clearly don't know about.

 

[hivebattleii]

Hmm.. how about two line method? little on the end and little on the other end.
I am getting sun beam core contact cooler ( which comes with TX-2) for Q9550.

 

[LAN_deRf_HA]

I use to be perfect at this, but lately I've had no luck. Last time it was so bad I couldn't get the heat sink off because there was so much paste it created a vacuum seal. I mean I swear I put the exact same amount on and get different results each time.

 

[sneekypeet]

Hmm.. how about two line method? little on the end and little on the other end.
I am getting sun beam core contact cooler ( which comes with TX-2) for Q9550.

isnt that a direct heat pipe cooler? With my Xig. I spread a bit on the cooler bottom as well to take up the gaps between the pipes and the "puck". Then applied paste to the CPU as well.

 

[hivebattleii]

isnt that a direct heat pipe cooler? With my Xig. I spread a bit on the cooler bottom as well to take up the gaps between the pipes and the "puck". Then applied paste to the CPU as well.

Yeah. it has coppers directly touching the CPU

Is it normal to spread it out and apply it all over the cpu before putting the heatsink?
Or should I just draw a line or two and apply heatsink right over it?

 

[sneekypeet]

I spread, but by no means am I saying it's a better method. just found I tended to use a bit less tim when I spread it myself!

 

[OzzmanFloyd120]

I spread out mine with a razorblade and never check, I'm afraid that if I do there might be small little air bubbles that get in there that will pevent proper conductivity.
As for NanoHead's comment, completely disregard that, he obviously has never used any other thermal compound than stock if he thinks there's no difference.

 

[nanohead]

I spread out mine with a razorblade and never check, I'm afraid that if I do there might be small little air bubbles that get in there that will pevent proper conductivity.
As for NanoHead's comment, completely disregard that, he obviously has never used any other thermal compound than stock if he thinks there's no difference.

Thanks for that sage wisdom

In my drawer at the moment, AS5, AS Ceramique, Thermalright ChillFactor, Zalman STG1, Balance Stars Gold, plus some Akasa and 3M tape.

I've used these and others over the years, and found almost ZERO difference. The heatsink design airflow make far more difference. I have probably 30 heatsinks and cooling gadgets in my office at the moment, and was one of the idiots who wasted almost $400 on a CoolIt Freezone (what a bad design, and expensive too)

I also have 25 years experience as an engineer in the computer industry, making things just like everyone here plays with every day.

In reality, the rise of TIM as the next big thing to argue about on forums is one of the funniest things in the world to most people who actually build this stuff. The guys at CoolIt said to me once "we put arctic silver 5 on because all the kids think its good, so it makes it easier to sell, but it really makes no difference".

I have NEVER EVER seen 10-15 degree reductions in temps. Not with processors, network ASICs, or power subsystems based on just changing heatsink compound. Just NOT GONNA happen. I could see that if you compared cardboard to a commercial thermal grease, but all this idiocy over TIM comparisons is just dumb. The guys that review this stuff actually put up .01 of a degree as a measurable difference and then put that on an excel bar chart and it looks like its a huge difference!!! Just like the reviewers that put up 1 FPS on a bar chart for a GPU review and state that the faster one "owns"..... just ignorant.

I would agree that the stock production line rolled tapes that are used on most production heatsinks (this is the dry grey clay like goopy stuff that is applied by machine to most heatsinks) is not ideal, and that using a more pliable material will yield better temp reduction assuming airflow and convection is identical, but in reality, the TIM thing is hilarious to those of us that actually understand this stuff.

It is almost as ludicrous as when all the 14 year old adolescents were telling everyone how many rails they needed in their power supplies a couple of years ago, and blaming the power supply for everything that was wrong with their systems. Silly and funny

 

[Lillebror]

Have you ever tried diamond tim? That stuff actualy make 10-15 degress reduction, if not more