Heatsink modding

[loko]

Hello Scooby,

My max temp of 55° is after streaming 1080p for an hour, your 46° is amazing.

Absolute temperature values cannot and should not be compared. Your room's mean temperature can be 18ºC, compared to 24ºC in another room. You may have placed your device close to the radiator, or other electronic devices which radiate heat. You may have placed your device close to other objects, which obstruct proper ventilation. You may have placed your device on a wooden surface, or on a cold marble surface which acts as a large heatsink. Different places in the same room may have a temperature difference more than 10ºC. The absolute temperature value of a device always depends on its external conditions. This is why the only way to evaluate a thermal modification is to compare before and after modification device temperature values under the same external conditions, also mentioning the actual room temperature, close to the device.

[Scooby-Doo]

[Scooby-Doo]

Hello Scooby,

My max temp of 55° is after streaming 1080p for an hour, your 46° is amazing.

Your mod looks very good. I expected, that you would have lower temps. Maybe I had only luck with my cpu. But 55 ° is absolutely in the specifications. The cores throttle their speed at temperatures over 65°. So you won´t loose performance in long time runs of the box. I think the Cpu isn´t the vulnerable part of the box. It´s rather the cheap power unit they put into the box threfore we need the cooling. Maybe the cooler has to work more because of the heat of the power unit.
This was the reason to put the cooling pad on the backside of the board. so it has contact to the alu case, wich can take away a part of the heat.
In all my computers I attach importance on cooling of the mosfet chips and a stabile and reliable power supply.

[locutus2000]

So I may sound stupid, but has anyone tried asking Geekbuying for compensation for the heat problem? Or have them come up with a thermal upgrade kit to be shipped to existing users? Perhaps someone can leverage their relation with Pipo for a solution for us early adopters?

[Pelliculart]

But 55 ° is absolutely in the specifications. The cores throttle their speed at temperatures over 65°. So you won´t loose performance in long time runs of the box.

Hi all and thank's for all Scooby, I like your tests ! ^^
I've high ° like 70-80 in video streaming Full HD. I don't care if the CPU is near 80-90° but I just want that cores don't throttle their speed until an higher degree.

Can I block secure throttle or jump the degree needed ? Can I just make a "minimum" frequency like 1 ghz ? (In bios or a windows program ?)

Thank's !

[Scooby-Doo]

Hi all and thank's for all Scooby, I like your tests ! ^^
I've high ° like 70-80 in video streaming Full HD. I don't care if the CPU is near 80-90° but I just want that cores don't throttle their speed until an higher degree.

Can I block secure throttle or jump the degree needed ? Can I just make a "minimum" frequency like 1 ghz ? (In bios or a windows program ?)

Thank's !

No. Look in the bios thread. The cpu is very limited. Silverlight and flash with DRM and HDCP costs alot of cpu. (look into the streaming threat). If you find a solution to liberate the cpu, than let me know. I couldn´t find.

[Pelliculart]

No. Look in the bios thread. The cpu is very limited. Silverlight and flash with DRM and HDCP costs alot of cpu. (look into the streaming threat). If you find a solution to liberate the cpu, than let me know. I couldn´t find.

Maybe I've find something !
In the pack driver Pipo X7 of Finless's post i've found an editable "Dptf" file (.inf) in the DPTF folder.
When you look into a text editor you can see some lines about the frequency, the processor, etc.
But i'm not good at windows driver and I don't know what to modify there. Can you check it ?
Thank's !


EDIT : I think i've found somethink, disabling/deleting Dptf driver can improve performance a lot, like said MichaelS81 here : http://forums.lenovo.com/t5/ThinkPad.../979927/page/9

[Scooby-Doo]

Maybe I've find something !
In the pack driver Pipo X7 of Finless's post i've found an editable "Dptf" file (.inf) in the DPTF folder.
When you look into a text editor you can see some lines about the frequency, the processor, etc.
But i'm not good at windows driver and I don't know what to modify there. Can you check it ?
Thank's !


EDIT : I think i've found somethink, disabling/deleting Dptf driver can improve performance a lot, like said MichaelS81 here : http://forums.lenovo.com/t5/ThinkPad.../979927/page/9

With this things I wouldn´t play around. Thats for software experienced guys. Maybe you find someone inthe bios thread, or you ask finless.

[gixie]

Thermal pads specs

Hi,
i need some clarification...
so you put one big thermal pad under the motherboard that has the thickness of 5mm
then you put thermal pads (1mm thickness) under the aluminum plate on the cpu and the voltage regulators and then heatsinks on the aluminum plates.
is this correct?
thanks

[Scooby-Doo]

Hi,
i need some clarification...
so you put one big thermal pad under the motherboard that has the thickness of 5mm
then you put thermal pads (1mm thickness) under the aluminum plate on the cpu and the voltage regulators and then heatsinks on the aluminum plates.
is this correct?
thanks

Yes.

[portezbie]

Two questions for you scooby:

1. What would happen if you just cut out a piece of thermal pad the exact size of the alu plate and put it between the plate and board instead of just cutting out little pieces? Would that be transfer heat to places it shouldn't go or something? Sorry, this is my first time learning about cooling.

2. The original alu plate has 2 little pads under it and you say added a third. I'm not sure if I can identify these parts of the motherboard. Anyway you could point them out on a picture or are they easy to identify?

Thanks as always, sorry for any dumb questions.

[loko]

2. The original alu plate has 2 little pads under it and you say added a third. I'm not sure if I can identify these parts of the motherboard. Anyway you could point them out on a picture or are they easy to identify?

http://www.freaktab.com/showthread.php?22473-Heatsink-modding&p=275898&viewfull=1#post275898

I have never got any answers from any user about the exact thickness of the two stock thermal pads, and whether they have exactly the same thickness or being somewhat different. To get a precise measurement, a caliper should be used, or even a ruler if being very careful.

Do the processor chip and the power unit chip seem to have the same total height? If yes, a single copper plate could be used for these two chips, or else a thermal pad would be necessary.

[portezbie]

I've seen those pictures, they don't address my question.

[Scooby-Doo]

Two questions for you scooby:

1. What would happen if you just cut out a piece of thermal pad the exact size of the alu plate and put it between the plate and board instead of just cutting out little pieces? Would that be transfer heat to places it shouldn't go or something? Sorry, this is my first time learning about cooling.

2. The original alu plate has 2 little pads under it and you say added a third. I'm not sure if I can identify these parts of the motherboard. Anyway you could point them out on a picture or are they easy to identify?

Thanks as always, sorry for any dumb questions.

Here a picture:



Maybe you can imagine.

[loko]

I've seen those pictures, they don't address my question.

Sorry, I thought you were talking about the back side thermal pad.

I put this one today:

http://www.amazon.de/Phobya-XT-W%C3%A4rmeleitpad-120x20x1mm-wei%C3%9F/dp/B0051C287K/ref=sr_1_3?ie=UTF8&qid=1423947281&sr=8-3&keywords=phobya+xt

I have read in various reviews that the newer Phobya XT 7W/mK is almost not sticky, very hard, and not compressible, while the older Phobya Ultra 5W/mK is quite sticky, soft and compressible.

When the gaps are not even, or there is a chance that the thermal pad may slip, Phobya Ultra is recommended by all experienced users for better results, while when the gaps are even and there is a lot of pressure, Phobya XT has obviously better results.

It seems that your setup is very effective, so there is no reason for you to go any further.

Anyway, my device has already been shipped and I have already ordered various modification components, so I will be able to check these situations myself.