Coollaboratory Metal Pad - overheating issue after a period

[redirect303]

Hello!
I'm honoured to join overclockers community. The main reason I joined is because I hope to find a solution for the issue that I have with my laptop and I thought this would be the best place to ask:

I replaced the thermal paste in my laptop (dell precision M6800) with Coollaboratory Metal Pad and it was all spring and cool for 1-2 weeks. Problems started to appear slowly, after that period when the overall temperature on my cpu started to increase and now is acting really strange when on maximum performance (when cpu stays always to 100%):

- when doing simple tasks like opening Internet Explorer or surfing the web, it has sudden thermal spikes to over 90 deg. Celsius, for a very short period.
- when the spike occurs, the fans go to full speed for few seconds and then they calm down.
- when on high stress, the cpu goes over 95 degrees

Mentions:

1. Yes, I burned-in the foil after I applied, by putting the laptop on the bed and firing OCCT until my cpu and gpu got aroun 70-80 degrees Celsius.
   
2. Also, as I mentioned above, all was perfect the first 1-2 weeks.
   
3. This is the second time when I apply the Metal Pad on my machine. First time it happened the same (was all perfect for more than 2 weeks, almost one month, when it went hot again)
   
4. A while ago I used XTU to play with my CPU (and I managed to get it into a sweet stable spot, by the way... but after a lot of trial) - could that damage my cpu? But if so, it would manifest immediately, not after weeks, no (also see pct. 4.) ?
   
5. Before applying, I cleaned the CPU with isopropyl alcohol and let it to dry
   
6. Also, before applying I sanded the copper plate on the heatsink with 5 micron lapping film until I cleaned all the oxidation and then I rubbed it with isopropyl alcohol
   
7. I applied Coollaboratory Metal Pad in my wife's laptop and there is no problem for months already (that one is way less performant, but still, the cpu stays cooler than it was originally)

I would welcome any input from whoever used Coollaboratory Metal Pad and could help me to rule out the problem.
My thought right now is that it may have happened to build some thin layer of something (oxidation, for example) that prevents a proper heat transfer from the cpu to the heatsink, despite the presence of the metal pad. Could that be a thing?


Many thanks!

 

[don256us]

What is the percentage of the isopropyl alcohol? The normal stuff is ~ 70% with the other 30% being skin softener/oil. Use 91-99% if you can get it. Our local Walmart has it for about $1.50 USD.


I'm a bit dubious of this metal plate cooling pad personally. For me, the tried and true method of thermal paste is what I use exclusively. There are many great brands. I do not use the liquid metal versions as they are too difficult to apply.

 

[redirect303]

What is the percentage of the isopropyl alcohol? The normal stuff is ~ 70% with the other 30% being skin softener/oil. Use 91-99% if you can get it. Our local Walmart has it for about $1.50 USD.

It was laboratory grade - 99.9%, so the contamination is totally excluded.

The metal pad is an alloy that melts at 60 deg. Celsius and is extremely thin foil. Like gold foil used by artists. And is supposed to melt and fill any minuscule gaps that are there. Plus, the thermal conductivity is much better than any thermal paste, because of the lack of additives. It's only metal.
Problem is that in my case it caused only problems and I can't tell why... Maybe someone on this forum used it and could give me an advice.

 

[||Console||]

Is it possible that it flowed where you don't want and causing shorting?

 

[redirect303]

Is it possible that it flowed where you don't want and causing shorting?

Wouldn't that cause BIG problems? I am using my laptop now as normal (actually responding to you reply from it)... I believe that a piece of metal on the resistors on the CPU would damage the CPU, wouldn't it?

 

[caddi daddi]

I have had to give up on all liquid metals, they have given up after a few months, migrated to areas I don't want them to go and caused me quite a few problems.
remove the liquid metal and try any other tim and see what happens.

 

[redirect303]

I have had to give up on all liquid metals, they have given up after a few months, migrated to areas I don't want them to go and caused me quite a few problems.
remove the liquid metal and try any other tim and see what happens.

What do you suggest would work best?

 

[Mr.Scott]

Isopropyl alcohol will not remove liquid metal. Only thing that removes it is lapping with a fine abrasive, like 2000 grit wet or dry. If you've applied more than once, that's why your temps suck.

 

[redirect303]

Isopropyl alcohol will not remove liquid metal. Only thing that removes it is lapping with a fine abrasive, like 2000 grit wet or dry. If you've applied more than once, that's why your temps suck.

@Mr.Scott, Please read the entire thread, your response doesn't address the real issue here. The issue is clearly described in the first post.
Thank you.

 

[Mr.Scott]

I read the entire thread. It doesn't change my statement. You screwed up the applications.

1. Yes, I burned-in the foil after I applied, by putting the laptop on the bed and firing OCCT until my cpu and gpu got aroun 70-80 degrees Celsius.
2. Also, as I mentioned above, all was perfect the first 1-2 weeks.
3. This is the second time when I apply the Metal Pad on my machine. First time it happened the same (was all perfect for more than 2 weeks, almost one month, when it went hot again)
4. Before applying, I cleaned the CPU with isopropyl alcohol and let it to dry
5. Also, before applying I sanded the copper plate on the heatsink with 5 micron lapping film until I cleaned all the oxidation and then I rubbed it with isopropyl alcohol

And then you come with this:

My thought right now is that it may have happened to build some thin layer of something (oxidation, for example) that prevents a proper heat transfer from the cpu to the heatsink, despite the presence of the metal pad. Could that be a thing?

It is multiple layers of liquid metal. Cleaning it off is one of the downfalls of using it. Only way is the way I told you.

You have to do both contact surfaces BTW. That means CPU and heatsink.

 

[maxfly]

WELCOME TO OCF Redirect303!

if your temps are getting that bad something has gone seriously wrong with the liquidmetal. i would definitely remove the liquidmetal asap and reseat your heat sink with a solid thermal paste. i use thermalgrizzly on my personal rigs for both cpu and gpu but for clients i use arctic mx-4 because of its longevity, reliability and cost effectiveness. it lasts forever (as long as 8 years) and doesnt dry out or separate. i buy the 20oz tube for $20 and it generally lasts a couple of years.
please post up some pictures of the cpu and heatsink before you reseat so we might see what has happened with your mount. this thread very well may help others that have the same or similiar issues.

easy guys, were all friends here.

 

[redirect303]

Thanks @maxfly for your suggestion for thermalgrizzly.

It seems I need to clarify something:
- it was NOT Liquid Metal! It was Metal PAD ( see here: http://www.coollaboratory.com/product/coollaboratory-liquid-metalpad/) This one is applied solid and it melts when brought to 60 degrees. Then, when one disassembles the heatsink from the CPU, it peels off nicely, like tinfoil (but much thinner). So it DOES not react like Liquid Metal, to get embedded in the metal.

- Also, the cleaning with Isopropyl Alcohol was to prepare the heatsink for the metal pad. Not an attempt to clean the metal from it! I cleaned the CPU and the copper from the heatsink with alcohol, thoroughly until there was no trace of any impurity that I could notice.
- @Mr.Scott - The way Metal Pad is supposed to be applied is: after you fit everything and assembly back your unit, you HAVE TO burn it in, in order for that foil to melt.

From reading the replies, I understand that many of you are very well acquainted with Liquid Metal application (which is not the case here), but it seems the metal pad is not that well known or used... hence this gap in communication.

 

[don256us]

To be sure, until your initial post, I had never heard of it. I had to google it and I did find it interesting. The company claims over 50 awards from various reviewers. However, given your troubles, I'd go back to the tried and true TIM (Thermal Interface Material) like artic silver 5 or better. A silicone based paste.

The fact of it all is, while the metal pad is probably better than TIM, the temperature difference between the pad and a good TIM is probably negligible. For me, I'd stick to something that I know works and not cutting edge.

- - - Updated - - -

Question: I know that this is a laptop. Did the heat sink have a thick thermal pad on it before you replaced it? Maybe there is an air gap the increases as in melts into place.

 

[redirect303]

Question: I know that this is a laptop. Did the heat sink have a thick thermal pad on it before you replaced it? Maybe there is an air gap the increases as in melts into place.

It had a gray (stock) thermal compound, but it didn't seem to be very thick applied.
Coollaboratory are selling their Metal Pads for Laptop with some copper foils (0.1mm and 0.2mm thick).
First time when I tried it, I put a 0.1mm foil. I made a sandwich: DIE - MetalPad - Copper - MettalPad - HeatSink. But I had similar issues after a while (almost one month). Second time I decided to apply the Metal Pad only, with no copper spacer. Now I have similar heating issues... it could be a simple coincidence, or it would be as you say some air build up).
As you mentioned it, I also thought it may be possible that during heating, some elements to thermally expand and to make a gap to appear that would allow air to interfere. The biggest issue is not the average temperature, but the spikes... when I start a demanding activity it jumps so high (even 98 deg. sometimes), even though is settles down a bit after, but having such a sudden jump kind of worries me.

 

[Mr.Scott]

It seems I need to clarify something:
- it was NOT Liquid Metal! It was Metal PAD ( see here: http://www.coollaboratory.com/product/coollaboratory-liquid-metalpad/) This one is applied solid and it melts when brought to 60 degrees. Then, when one disassembles the heatsink from the CPU, it peels off nicely, like tinfoil (but much thinner). So it DOES not react like Liquid Metal, to get embedded in the metal.

- Also, the cleaning with Isopropyl Alcohol was to prepare the heatsink for the metal pad. Not an attempt to clean the metal from it! I cleaned the CPU and the copper from the heatsink with alcohol, thoroughly until there was no trace of any impurity that I could notice.
- @Mr.Scott - The way Metal Pad is supposed to be applied is: after you fit everything and assembly back your unit, you HAVE TO burn it in, in order for that foil to melt.

From reading the replies, I understand that many of you are very well acquainted with Liquid Metal application (which is not the case here), but it seems the metal pad is not that well known or used... hence this gap in communication.

It's liquid metal in pad form. It's the same thing. Even your link says "Liquid Metal Pad"
Same chemical, same reaction.

I'm wasting my breath. oy

 

[redirect303]

It's liquid metal in pad form. It's the same thing. Even your link says "Liquid Metal Pad"
Same chemical, same reaction.

I'm wasting my breath. oy

Thank you @Mr.Scott for the effort to convince me about something that is clearly you only assume, without an effort to study the issue.
I started this thread hoping I will find some advice from people who used Metal Pad, who can give me an advice based on facts and experience. Please read the facts below which clearly distinguish the physical and chemical properties of the two materials:
---
-Liquid Metal is a gallium based alloy (that's why is liquid at room temperature).
-Liquid Metal compound will "embed" itself into the superficial layer of the metal (CPU IHS or Heatsink), therefore it needs to be removed either with chemical means (chemically disolved) or mechanical, by "sanding it off".
-Liquid Metal forms a brittle alloy with aluminum (because of Gallium), that is why it should NOT touch any aluminum parts
-Liquid Metal is applied liquid, stays liquid and tends to go everywhere if not taken proper care.

-MetalPad is indium/bismuth based alloy (they say it has copper and tin too)
-Metal Pad DOES NOT embed into the metallic surfaces it gets in contact with. It actually looks like a thin foil of solid metal, which can be peeled off. It can leave some traces behind, where it breaks (as it's very thin), but even those can be scrapped off, without much force (i used a plastic spatula).
-Metal Pad DOES NOT react with aluminum, because it doesn't contain gallium
-Metal Pad is solid when under 60 degrees celsius. I am not sure how it behaves after the first melting, if it doesn't get any metallic ions from the heatsink which could its alloy properties (to increase its melting temp). I couldn't find any info regarding this.
---
*On the bottom of the page for each product on coollaboratory's page you can find links for manual and technical sheets... more info there.

So, NOT same chemical, NOT same reaction.
What is the same is: both are metals, so their heat conductivity is a feature of metals in general.

 

[Lochekey]

What type of mounting hardware is used to attach the cooler to the cpu? Is it possible that you are not getting a secure mount that is allowing the mount of the pad to crack and not transfer heat well. Traditional paste material has a little "flex" to it and can be a little more forgiving about the mount.

 

[redirect303]

The heatsink is fastened with 4 screws on what appears to be an X-bracket on the other side of the motherboard, under the cpu.
The screws are all with tension springs, so there is no way to over tighten them. Once the screw reached its end, then only the spring will do any pressure.
Do you believe that due to paste flexibility and layer thickness the contact was more uniform?

 

[Lochekey]

It is not that it is necessarily more uniform, it is more that if there is movement in the heatsink the paste can flex and survive the shock without physically separating from the block where as the metal pad will bit flex as well and will as such have a small physical gap form.


This would possibly be exasperated if the shootings in the mount are getting weak.


Please note this is all speculation on my part and by no means is it the guaranteed problem.

 

[don256us]

@ OP. I think that the clear answer is that we don't know. This is an unusual product that I don't think any of us have seen. We are trying to give you our best guesses but we do not have any experience with this product. I wish we did and I do appreciate that you've shared what you have to date. This is an interesting product.

What has the manufacture said about this? I'm sure they would have far more insight than we. If you have or get an answer from them, would you please let me know? I am truly interested in a positive outcome.