*PICS* Shin Etsu X23 Repaste Sager 8170

Good Morning
I just wanted to share my Shin Etsu X23 experience with yall.
These temps are only a few hours after the repaste. After 2 heat cyces the temps have dropped a few degrees.
I hope they drop more. If not i will try IC7 Diamond.
I currently use a NZXT CRYO LX Laptop Cooler and these temps are without ever using the FN + 1 key.
Ambient Temp- 25C/75-76 Fahrenheit
GPU Idle- 52C
GPU Load 3DMark Vantage- 74-78C
CPU Idle- 37-38C
BFBC2- 77-79C

EDIT!!!! My GPU sits at 1.00v in 2D mode instead of .085v. This is why my idle temps are so HIGH. When I run my laptop off battery my GPU Idle temps drop to 39-40C
Here are some pics!

 

Those temps are great. I don't understand why you expect any lower.

 

A bunch of the other Sager 8170 owners are seeing 38-39C idle temps
My load temps are good but my idle temps are 13-14 degrees hotter than most others

 

yup... 52C pretty high for idle.
Those are my temps for normal use (browse, office etc).

 

Damn, i want my GPU idle to be lower like everyone else's

Here is my temps
I just ordered a tube of ICD7.
I hope my temps drop.

 

To be honest I feel like shin etsu is overrated by lots of people. My corsair h100 come in with stock shin etsu and I have to say the result suck balls, however I'm not sure if it's the shin etsu or I have too high of expectation on h100, I will somehow repaste with some AS5 and see the result. Also i'm considering of repasting my np8170 to as5 from stock when i receive it later on today .

 

Part of the problem is that you pre-spread the paste instead of just letting the heatsink spread it out with the clamping pressure... there are tests confirming the fact that the spread method leaves air bubbles in some areas.

 

I think your problem might be that the GPU is always @ 1.00V ... no idea why.
As you can see mine is at 0.85V in normal 2D use.

 

@ Handsomerob, I cannot see your first post pics. Can you adjust it some?
_

 

I don't know if application methods are that different for laptops, compared to desktops, but that's definitely not how I would have done it. Seems like you're using way too much. The way I would do it is squeeze a small line along the chip and let the heatsink do the spreading.

 

We have used the recommended way below. This is how the factory at Sager has done it....


_

 

What do you mean adjust? Can you not see any of the pics?

 

One of the "most" important things that needs to be done in any reapplication of a TC is first removing the old TC completely from the CPU,GPU and heatsinks.

A small amount of 100% Isopropyl alcohol is good solvent to preform the old TC removal.

In applying the new TC just use enough to fill in the pores in the metal that forces the poor conducting air out creating maximum contact between the chip and heatsink.

Too much TC is a bad thing.

Tuniq-TX2 and TX3 work very well for me but not the new TX4.

A small amount of TC can be placed on the chips and heatsinks and worked in and around with a finger covered with plastic wrap.

 

As you can see in the pics, my CPU and GPU heatsinks are as clean as you can get them. I used 99% isopropyl alcohol and Arctic Clean.
I have repasted quite a few heatsinks in my time. But Shin Etsu is a new one for me and after lots of reading online it seems X23 works best spread out on the heatspreaders using a razor blade.
Either way my temps suck ballz and ICD7 is in the mail.
I'll post my results in a few days
I will be doing the pea method as LPC says

 

How Thermal Compound Spreads - YouTube HEY, perhaps this video might help you. Shows you proof that the pre-spreading method can lead to air bubbles.

EDIT: My GPU idle temps are at 40-45C but my ambient temps are around 35C.

 

i will try ICD7 and repost my results

 

Also, if you haven't seen it, Innovation Cooling has the application instructions posted to their site if you need it:

Application

 

I used the spread method with the IC7 and my temps are around 34-36 degrees idle. I tried the pressure method, but IC7 is too thick to reliably apply it that way.

 

You aren't running a second screen are you? If you are then it would explain your higher idle temps as the gfx card won't downclock.

 

No im not running a second screen.
I just realized i didnt tighten down the heatsinks in the order i was suppose to in the service manual.
I wonder if maybe thats why my idle temps are so high. Load is only about 79-82C, which is on par with everyone else. Its just my crazy high idle.
Just sitting hee typing this message and its at 53C

 

How did you spread it?
I wish there was a pressure mod. I remember doing a presuure mod on my M17X Alienware and i dropped 16 degrees

 

I spread my IC7 with a latex glove. The dot method has never reliably worked for me.

 

This is a good point. Handsomerob, why is your 485m not at 0.85V when idle/in 2D? That could explain the higher idle temps

 

Why is my 485M at 1.00V?? How can I troubleshoot this?
Thats probably why its so hot. What do i do?

 

Hardwaresecrets does a monthly thermal compound test.They add about 5 new TCs to their roundup every month.

The Shin-Etsu MicroSi was tested in the May Roundup.

Thermal Compound Roundup - September 2011 | Hardware Secrets

 

When I unplug my laptop and use the battery my Voltage of my GPU goes to 0.85v and my idle temp is about 37-40C and when I plug my computer back in it pumps the Voltage up to 1.00v and my idle temp is about 56-60C

Why isnt my computer dropping the voltage of the GPU to 0.85v when in 2d mode? What do i do?

 

My spreader of choice is a Also, follow Kevins advice and use a latex glove while do it so you don't get any oils mixed in with the paste. It might take some practice but shoot for a nice thin layer of IC7 over the entire chip. Even with air bubbles, my temps were still much lower than the pressure method.

 

I think I have found a solution to your problem .
Open the Nvidia Control Panel and go to Manage 3D Settings. Once there find the option "Power Management Mode" and put it on "Adaptive".

If it is on "Prefer maximum performance" than the video card will be @ 3d frequencies (1.0V) even in 2D.

Let me know if this worked

 

I tried this and it still does not change from 1.0v to 0.85v
What bios does your 485m have? Also which drivers are you using?
I'm using the newest Nvidia drivers

 

I'm using Nvidia BETA 285.38 driver.
As for the bios please see in the attached pic.

Mine is a Sager 8150...

Anyway, your issue is definitely because of the 1.0V in 2D.
I tested and put i on "Prefer maximum performance" an got the same temps in 2D as you ~60C. All this because it was running @1.0V.

 

Thanks for the help. My problem is now fixed.

 

So, Rob, does that mean the issue is not caused by the TIM ?
Most issues regarding TIM concerns the application, not the Make of the material. The dot/pea method seems best but some TIM have less viscosity than others, so when using this method, it is best to warm the tube of TIM in hot water for 5 Mins. prior to application.
Other note of concern is that with most Laptops, the heatsinks uses TIM pads as an interface therefore there will be a small gap and this gap needs to be plugged either by slightly bending the heatsink or use a thin copper-shim, I always use the latter.

 

I agree. The dot/pressure method is what I've always used and have always gotten great results. I too warm mine up a bit prior to application but I personally stay away from water I just rub it back and forth in my hands for a minute or two.

 

The problem was the voltage. It was always running at 1.0v. It never kicked down to 0.85v while in 2D mode. I opened up the Nvidia Control Panel and put the setting at "Adaptive" instead of Prefer Maximjm Performance. Now my GPU idles where it should be. I just did the "pea" method with ICD7 and my temps are awesome. I now idle at 37-38C(GPU) Load 75-78C(GPU) and my CPU idles at 37-40C. I'm also overclocked to 625/1550

 

what room temp u have??

 

Room temp is about 72-74F
And I put the ICD7 in a warm toaster oven before I applied to the heatspreader. It worked great. Just keep an eye on it so it doesn't melt the plastic. I'm very happy with the load temps. It never ever gets above 78C even after 4 hours of BFBC2.

 

Hey thanks for posting this issue. I will be expecting a similar laptop with a gtx485 within the next couple of weeks, and this will be first on my checklist of things to check. I have 4 tubes of the Shin Etsu but I'm going to see how the stock paste does before I reapply.

 

dont bother with the shin etsu
ICD7 is the shiznit

 

Firstly, in the OP Title, it doesn't state which Shin-Etsu TIM, is it X23 7762 0r X23 7783D ?

I have to disagree with the opinion of not bother with the Shin-Etsu as I find the X23 7783D to work best, for me. Much of the time it is the application and not the TIM. I have tried all the Pastes and am disappointed with the results, then returned to the X23 7783D.
Only just recently, as I don't have enough remaining Shin-Etsu TIM and due to the availability of the product, I purchased the Phobya HeGrease Extreme because it was touted as being extremely good. My result was a higher idle temperature of + 7. Along with Arctic Cooling MX-3 etc etc etc, it is all advertising to sell the products, if AC MX-3 is sooo goood, then why did AC sooo quickly replace it with AC MX-4 ? the answer is quite elementary.
Most reviewers contradict each other, some goes so far as to introduce Mayonnaise into the equation in which case, how can one see the integrity of these comparisons ?

I surmise that all products have a finite limitation and the Shin-Etsu X23-7783D is near the limit as far as TIM pastes are concerned. There are rumors that Shin-Etzu will have a new paste coming out soon called X23-7921-5. Obviously it will also be touted as better just to sell the product as consumers will always seek a new & better item, but can it better a product that is at the peak of its' limitation ?

Another scenario of human misjudgement is people lapping their CPU to achieve better heat transfer, it is one of the most stupidest things one can do. Firstly , it voids all warranty, then it damages the CPU.
I would put all my money on the people who invests billions of dollars to develop these processors that they have done their best (not to put hindrance of heat transfer) for their products, rather on a snotty-nosed kid playing modifications in the basement trying to further improve that component, but that story belongs to a different thread.............

There is one major problem with Shin-Etsu X23-7783D, it is darn difficult trying to find a retailer who stocks this product.

NB: I too tried ICD7, at the stage when my usual TIM was AS5 (200hrs) before I switched to X23-7783D (no curing)

 

Yup, I've been using 7783D for a couple years and have been really happy with results. In fact it holds up my builds when I run out. The scratches the ic7 makes was my original issue with it.

 

I used X23 7783D. I just swapped out pastes and now I'm using ICD7. My temps are about the same as the Shin Etsu if not better. The X23 7783D is easy to find. There are a bunch of peeps selling it on eBay for about $8 shipped

 

Just finished reapplying the Phobya HeGrease Extreme because the previous temperature reading of +7 in comparison to the X23-7783 is too great a difference to be realistic. Result is fluctuating between +1 to +2 which seems more correct. It's a matter of trust and I would like to return to using my previous TIM.

I already checked the ebay seller, atm it's US$6.99 with $1.99 shipping in US, but the shipping to my location is US$7.99, which make this 1g of TIM to be valued at $14.98, seems very expensive. Rather than making this purchase, I'm waiting for the S. E. X23-7921-5 to arrive at retail, I believe that a certain Overclocking forum may be having a group buy when it is available.

Edit: How does one tell if they have received the correct Shin Etsu product as there are no product numbers on the syringes ?
G751 = Syringe + Blue cap
X23-7783 = Syringe + Red cap
X23-7762 = Syringe + Red cap ??????????

ATM the only ebay seller of the X23-7783 has a stock photo of the X23-7762 in the Listing ???????

 

Okay guys, i have just recieved my new Radeon 6990M in the mail and a tube of ICD7, what method appears to give the lowest temps

1. DOT apply it and let the cooler spread it out
2. Pre-spread it out in a thin layer

Even if there is a only 1-3C in difference between these 2 used methods, i still want the absolutely lowest temp call me an enthusiast

 

Our technicians mostly use the dot method, though we have had special requests before for the spread technique. I've found from personal experience that the dot works best.

 

Just want to share this link:
2011 Thermal Compound Roundup – Results Compilation | Skinnee Labs

I just ordered Phobya HeGrease Extreme and hopefully i can test them in my upcoming new P170HM on November (i7 2670 and 6990m).

 

All these TIM Reviews/Comparisons, they all have their own results. Most Sites are either bias due to gratuitous freebies from manufacturers for a little better result or their application of the TIM are sometimes flawed.
Another words, DO NOT believe/trust these tests/reviews.
Most of the best TIM produces similar results but it is the application which determines the best result. Some Sites put Mayonnaise, Cheese, Ketchup into the Reviews equation, so how can you believe the integrity of these people ?
If you see someone touting that theirs is The Best, it's all cow-poo, there is no best TIM.