Applying Shin-Etsu

So, while is seems like there are a million different ways to apply a million different compounds, I'll probably be just fine if I use a little MX-3, applied with my finger in a baggy? Credit card it smooth, or just even it out with my finger? And then just press down on the heat sink when I put it back on?

Is this basically what many of you guys would consider a decent, perhaps even the best, way of application?

 

A card would work better for smoothing it. Just never user your bare finger; the oils will ruin the compound.

 

Some prefer to smooth it out with a card or joker playing card, while my self and other believe the dot and pressure method are better. How ever the pressure method must be adapted as the heat sink alone will not provide enough pressure. I would suggest, since everyone has their own method is to do what you are comfortable with.

 

so, since the heatsink in a laptop does not supply the needed pressure to properly spread out the paste...is it ok to push down on it to provide a little extra squishing? would that endanger the mobo the cpu/gpu? i know it was mentioned earlier in this thread that newer cpu's have a protective layer over the top to prevent the die from being scratched...so maybe just the mobo then.

does ICD7 really dry out? thats crazy, since its suppose to be "the best" i guess ill have to look into x23 then. or would you suggest AS5? i suppose it doesnt really matter as long as its done right...as Soviet has said numerous times.

Edit: I found this link comparing ICD7 and X23, thought it might be helpful:
http://www.overclock.net/cooling-experiments/453054-ic-diamond-7-vs-shin-etsu.html

 

oops sorry i started reading at page 3 or something, missed the video on the first page

my heatsink is secured with 4 large screws if that wont provide enough pressure i dont know what will... when i did remove the heatsink to switch cpus the paste was nicely spread out and my cpu temps have always been minimal. pea or line method ftw! anyone using the spread method is wasting his time imo

 

ICD7's composition, just like X23 and the vast majority of other pastes on the market, uses an aluminium oxide silicone foundation. It will eventually dry up over time and lose it's ability to transfer heat. This usually takes years of thermal cycling to break in. But even then, because the thermal paste is still there, the increase in temperature is probably not going to be more than 5*C, which is really not much. This is why thermal paste manufacturers say that thermal paste can last as long as a decade or whatever length of time they rate it for. But the maximum effective service time for thermal paste until needing to replace it is about a year. It's just like replacing the oil on your car. You can't just buy the best thermal paste on the market and set it and forget it. For me, I usually replace my paste on a seasonal basis, mainly because it gives me something to do and it's just a typical part of my maintenance routine.

 

good to know!

new question: Where can I buy X23 in the states? is there anywhere local? i dont think Tiger Direct, Newegg or Best Buy carries it

 

http://www.google.com/webhp?hl=en

 

good. thats...good.

 

frozencpu has X23 as well as MX-3 and some other harder-to-find stuff.

 

I got my X23 and MX-3 from Frozen. I am going to swap to MX-3 and see if my GPU temps improve any over OCZ Freeze. I couldn't get good temps using X23 on the GPU (CPU is great) so we'll see which of the two is better.

 

They should be pretty similar.

The way you apply it (and how well you do) will determine which is better

 

The X23 also needs time to cure, if its anything like IC it could be any were up to 200 hours.

 

Another reason I like MX-3 better (no cure time )

 

There is that depends if you want the extra 2-3*C...

 

http://forum.notebookreview.com/showpost.php?p=5441586&postcount=507
http://forum.notebookreview.com/showpost.php?p=5619986&postcount=1455

 

Thanks SS so general rule of thumb is after a good day of using the Notebook the temps you see are the temps you going to get.

 

Yes, usually after a day of letting the thermal paste sit there while the notebook is operating is enough. Anything longer than that will yield fractional results.

 

Hi everyone

I am new here and was considering to replace the stock paste of my g51 CPU and GPU recently.I came across with many different brands like AS5,IC D7,MX-3.......Yet I would like to give it a try on Shin Etsu.

But still for the 2 choices:Shin-Etsu MicroSi G751 and Shin-Etsu MicroSi X-23 7783D-which one is better,ignoring price?or they perform more or less the same

Thank you for any answers.

 

X-23 is the one to get.

 

Ye totally agree !

 

So X23-7783D is better than G751 then ?Cheers guys I am going to get one now!

 

Ye its a better revision of the old paste, and meant to be easier to apply than its predecessor can you believe

 

I just came across this combination technique while browsing another forum, let me know what you guys think:

First, put a very small dab at the center... spread this around using your finger in an inside-out plastic bag (take care to have your finger very tight against the plastic bag... this is to prevent aren't any creases in the plastic bag which could cause uneven distribution/tiny air pockets)... you want to have a fairly thin layer - if it doesn't seem thin enough, wipe off the excess TIM from your finger/plastic bag with a cloth and alcohol then respread (this will lift off some of the TIM, thinning the layer)

After you have this nice layer, apply the thin line down the center (parallel the notch for quads - perpendicular for dual cores)then simply attach your HSF...If both surfaces are flat, the spread layer will ensure full surface contact while the line ensures no air bubbles and good contact at the main heat source on the IHS

 

Never tired it so cant comment, I think I have created a nervous bunch of people by making this thread. I think that either the dot or the spread method can work fine depending on the person that is applying it.

Earl, you could try that and post back and let me know what temps you get

 

bah that required initiative on my part but i should get my x23 in about a week so ill let you know.

 

Actually, your method is VERY similar to mine that I came up with one night laying in bed (dead serious.) Instead of using a plastic bag I used latex gloves. Needless to say I did exactly what you said except one big problem.. When you spread the X23 it works for about 5 seconds then starts to clump up and you get some areas that have a clump of thermal paste and others that have none. I tried "filling in" the spaces where the thermal paste was lifted up off onto my glove and applying the heatsink and my temps were just god awful on both cpu and gpu. Using 2 small dots on my CPU netted good results and for my GPU I never could get the X23 to perform well at all. I ended up switching to OCZ Freeze and I dropped temps overall 20c (obviously from how I was applying) from when I first tried the X23 on the 5870 and then ultimately 10c over the X23 after I tried 4x and gave up.

 

Proper application of X23 is paramount. Using a plastic bag or gloves with X23 will end up with a sloppy application. The "bag" method is not meant for X23. If you cannot properly spread X23, then just use MX-2 or another low viscosity paste. OCZ Freeze is also one of them. It's really that simple.

 

And what temps are you getting know both at idle and under load

 

Me I am still trying to figure out what I want to use.

I was set on ICD, then Shin-Etsu, now I am looking at MX3. I know any would work just fine. Just the Engineer in me research, think, research some more think more, become more undecided as you learn more. Then aw sh*t just pick one and be happy.

 

lol ye I hear ya! im sure any you suggested would be fine.

 

hahahaha i feel you, i guess all engineers are the same you should see me shopping for a pair of shoes (anything that costs more than 20 dollars really), i drive my girlfriend crazy

 

OMG! there is someone else out there like me. I feel so much better now. It's crazy I do the same, investigate, then hunt down the best price. My wife get's a kick out of it.

 

I already posted it somewhere but I'll post it here again.

80-way Thermal Interface Material Performance Test

 

so since there is such an emphasis on applying the paste correctly, could someone that feels confident in their ability to apply X23 post a video of said activity?

 

I got some MX-3 and today, before my8760 gets here, I decided to use my Vaio as a guinea pig. First of all, what a b*tch to get that thing apart. But the difference in temperature is fantastic. I've lowered my CPU temps by 14* C. I spread it on with a playing card.

I think it's more the way Sony put their paste on sucked, than the MX-3 being a miracle paste. Still though, it was very easy to apply. I also bought a solution (ArctiClean) to clean the old paste off. I dunno how it works with regular alcohol, but the ArctiClean melted the old paste right off.

I will report results when I do the 8760 as well. But I do want to thank Fluffy, Soviet Sunrise, Sleey, et al for their comments on how to.

 

@User Name: Nice to hear it 14*C sounds good! know think what temps you will have when coupled with the amazing cooling of the Clevo.

Gppd luck and let me know how it goes

 

Just wondering if there is a good alternative to a professional cleaning solution that can be found...hm maybe in the kitchen or somewhere?

 

Standard isopropyl alcohol works as well, although you'll probably have to put in a little more elbow grease to get all of the old paste. I think 70%-90% is recommended.

 

Ye thats what I used came of a treat didnt melt away like it does with some cleaners but it wasnt a hassle.

 

Is it important to remove the GPU and CPU prior to cleaning so as to avoid possibility of the solvent from running down into the mobo?

I heard on desktops that extra solvent can drip down into the pegs that connect it to the mobo and cause random blue screens.

 

You shouldnt be using that much! Its only enough to wet the surface, in all thinks thermal paste less is more You dont need to remove anything.

 

Even so, if the cleaner is isopropyl alcohol, it should dry very quickly by itself anyway.

 

does it matter if you use 70% vs 90%?

 

The more concentrated the quicker it evaporates. And it also contains less water which will harm your lappy. Minimum 70%

 

so if i have 70% should i bother going to the store to try to find 90% ?

 

No that should be fine just ont over do it.

 

Curious, are you guys letting the thermal compound cure a little bit before putting the heatsink on? I found tonight letting it sit there for about 5 min before applying the GPU heatsink netted the best temps. I also started wondering if it is best to apply the paste, apply the heatsink and let it sit for an hour or so before starting it up. This would allow the paste to "bond" to the heatsink a little better. I've been swapping paste around so much I just didn't want to take the time to let it sit in between.

I have been doing a small shootout between X23, OCZ Freeze, MX-3 and Artic Silver 5 in the hunt for the best temps and so far the OCZ Freeze surprisingly is the best for me by about 4 degrees. I tested each paste 4x using different methods and every time the OCZ Freeze worked the best and all 4 times the temps were within 1-2c of each other. The other pastes are not near as forgiving and really depend on how much you use and the application method. I found for the MX-3 it took using a little more paste to get good temps where as the X23 required very little. AS5 was in between and is closer to OCZ Freeze in that it is forgiving in how you apply it.

I did the same 4 application methods each time and made sure to be using similar amounts (applied one paste, re-assembled and booted into Win7 and let it run for about an hr.) In the end the "dot in the center" and the spread method using something bendable/flexible overall worked best for me with the GPU. For the CPU the dot in the center and using the heatsink for pressure worked best with X23.

I'd love to see someone make a video of applying X23 to the 5870 core. My end results were OCZ Freeze on GPU and X23 on CPU after about 30 applications between the various pastes over the course of a week

P.S. - Bring back Soviet

 

update on my x23 project: I just applied it about 40 min ago (and somehow my computer survived after my girlfriend dropped the screwdriver on top of the bare cpu)

temps at load with stock paste:

CPU: 55-62C
GPU: 73-86C

temps at load with X23, no cure time:

CPU: 59-63
GPU: 68-76C

also I believe that I did a god awful job applying the compound, so if i redo it tomorrow this is the baseline i guess.
anyone have a steady hand? i am taking applications for a new assistant.

also I feel I should add that by "at load" I mean I am running dual monitors, one with assassins creed at max settings, the other playing youtube and surfing the web.

 

Earl - Def redo. X23 doesn't take but about 2-3hrs max to cure so if your temps haven't gone down I'd redo it. From personal experience you don't get but maybe 1-2C anyways after it cures. I have the same proc as you and mine runs 42-45c at idle and GPU is usually in the 55c range.

If your temps are under load then I'd say you may of did a good job. Post your idle temps.