The MASM Forum

General => The Soap Box => Hardware Corner => Topic started by: hutch-- on July 26, 2020, 12:12:19 PM

Title: The box, finished !
Post by: hutch-- on July 26, 2020, 12:12:19 PM
This has been a reasonably tedious box to build, board was a pain to get going with having to flash the BIOS. Apart from a 4 hole USB face plate its finished, working correctly and performs well. A legacy device, its pointed at being a work horse, not a show pony. The can is a bit smaller than I like but managed to shoehorn 5 HDDs and one SSD into it. Future mods when I have some spare cash is another 64 gig of memory and perhaps an NVMe drive for where I need fast loading.

For your pleasure.  :tongue:

(http://masm32.com/private/TheBoard.jpg)

(http://masm32.com/private/TheBox.jpg)
Title: Re: The box, finished !
Post by: LiaoMi on July 26, 2020, 06:59:27 PM
Hi Hutch,

congratulations on the completion of the project! What kind of thermal paste did you use?
Title: Re: The box, finished !
Post by: hutch-- on July 26, 2020, 07:15:06 PM
The latest version of Arctic Silver. The old (10 years old) version always worked well so when it eventually ran out I bought a couple of syringes of the new stuff. There is a trick though, instead of a tiny dab in the middle, use enough to coat both processor and heat sink then put them together with a little bit of wiggle to seat it properly then screw it all down. I still use my finger.  :biggrin:
Title: Re: The box, finished !
Post by: LiaoMi on July 26, 2020, 11:21:59 PM
I used ic diamond https://www.amazon.com/IC-Diamond-24-Carat-Compound/dp/B0042IBAOG (https://www.amazon.com/IC-Diamond-24-Carat-Compound/dp/B0042IBAOG) at my own risk. Experts say it can corrode the surface.

Quote
Contains 92% Purified synthetic diamond micronized Diamond, a natural thermal superconductor with a conductivity of 2,000-2,500 W/mK compared to 406-429 W/mK for pure silver
Will not pump or bake out when properly used and is neither Non capacitive or electrically conductive. C of C available on Request
Title: Re: The box, finished !
Post by: hutch-- on July 26, 2020, 11:58:58 PM
I have seen some of the guys on Youtube use what they call liquid metal but it is suppose to be corrosive. I tend to take a safe route that I am familiar with and the Arctic Silver has been super reliable for years. I have taken off a heat sink that I put on 10 years ago and it has not baked hard, is easy to get off and non corrosive.

I am clocking these 3.3 gig i7 5820s at 3.8 and the temperature is very low with a decent cooler so I don't have much to gain by using a technique that I am not familiar with. Have you done any comparative testing with what you are using ? On highly over clocked processors I have seen results of 5 to 10c reduction but it often is related to de-lidding a processor first as the junction between the core and the lid is critical.

I have even seen for sale pure copper replacement lids and I gather they are more efficient. I had a look at the link and it looks like interesting stuff, would like to see some tests done on CPUs.
Title: Re: The box, finished !
Post by: LiaoMi on July 27, 2020, 01:11:45 AM
I have seen some of the guys on Youtube use what they call liquid metal but it is suppose to be corrosive. I tend to take a safe route that I am familiar with and the Arctic Silver has been super reliable for years. I have taken off a heat sink that I put on 10 years ago and it has not baked hard, is easy to get off and non corrosive.

I am clocking these 3.3 gig i7 5820s at 3.8 and the temperature is very low with a decent cooler so I don't have much to gain by using a technique that I am not familiar with. Have you done any comparative testing with what you are using ? On highly over clocked processors I have seen results of 5 to 10c reduction but it often is related to de-lidding a processor first as the junction between the core and the lid is critical.

I have even seen for sale pure copper replacement lids and I gather they are more efficient. I had a look at the link and it looks like interesting stuff, would like to see some tests done on CPUs.

The difference in base temperature is 4-6 degrees, under load, the temperature is 7 - 11 degrees lower. The same was promised in the technical documentation. The fan always runs at the same speed, and the outgoing air became hotter, it scared me at first. I would say that the cooling works smoothly. I applied the thermal paste very carefully, because based on reviews, the erosion was not caused by the paste itself, but by strong surface friction. Therefore, I did everything using the 5 drops technique  :biggrin:
Title: Re: The box, finished !
Post by: hutch-- on July 27, 2020, 01:53:38 AM
It sounds like something you would use inside the processor once the lid is removed. You have two heat transfers, die to lid and lid to cooler and it seems that different processors transfer the heat at the first boundary poorly. I have an i7 4770k that runs at 3.8 OK and the temperature is not all that high but go the next notch at 4.1 gig and it runs at idle OK but thrash it and it crashes without getting that much hotter.

It looks like useful stuff, if its not too hard to get here in OZ I could be tempted to try it. These socket 2011-3 processors are physically large and you have a big dissipation area given that they run at 140 watts.
Title: Re: The box, finished !
Post by: LiaoMi on July 27, 2020, 04:08:46 AM
This is how it looks from the inside...

(https://www.eteknix.com/wp-content/uploads/2014/06/MSIGT70_DomPro_Cooling.jpg)

(https://eteknix-eteknixltd.netdna-ssl.com/wp-content/uploads/2014/06/MSIGT70_DomPro_Heatsinks.jpg)

(https://eteknix-eteknixltd.netdna-ssl.com/wp-content/uploads/2014/06/MSIGT70_DomPro_CPUClose.jpg)

https://www.eteknix.com/msi-gt70-2pe-dominator-pro-gaming-notebook-review/3/ (https://www.eteknix.com/msi-gt70-2pe-dominator-pro-gaming-notebook-review/3/)

Thermal Throttling in the original assembly - https://i.imgur.com/y0rtWAJ.png (https://i.imgur.com/y0rtWAJ.png)

Quote
You have two heat transfers, die to lid and lid to cooler and it seems that different processors transfer the heat at the first boundary poorly.
Exactly!
Title: Re: The box, finished !
Post by: hutch-- on July 27, 2020, 11:10:57 AM
Its a reasonably difficult task to properly cool a laptop and the cooling pipes in the photos show that its a high end laptop. The 82c is clearly too high so if you can get that down it will certainly improve performance. I don't think there is much you can do to improve ventilation in such a small form device so it will be interesting to see what reduction you can get with the thermal paste you are using.

In an enclosed space an ordinary SATA SSD will run a lot cooler than an NVMe drive and I don't know if there is much more you can do to get the temperature down. I have seen some form of coolers that you sit the laptop on but I don't know how well they work.
Title: Re: The box, finished !
Post by: daydreamer on July 28, 2020, 03:59:57 AM
congrats on new box Hutch :thumbsup: