those of you with Liquid Metal TIM, did your temps increase after 2 years?

[Space Lynx]

poll

 

[R-T-B]

Not I, said the frog.

Copper heatsinks need not apply, though. NIckel-plated only.

 

[Space Lynx]

Not I, said the frog.

Copper heatsinks need not apply, though. NIckel-plated only.

I didn't know that. Noted, if I ever do liquid metal again it will be nickel only

 

[Frogger]

Rock steady from the day i did the delid 5+ years...

 

[infrared]

I have noticed there's a 5c discrepancy on my 6700K, but tbh I think it was always like that, probably just slight variance on the sensors. Temps are still excellent. That's with bare die onto a copper block, this pc's been together over 3 years now and I don't plan on removing the block any time soon. Voted no.

 

[MrGenius]

On copper it may need reapplied. Not even always. Other than that it lasts forever.

 

[AltCapwn]

Not I, said the frog.

Copper heatsinks need not apply, though. NIckel-plated only.

The wisdom of the frog as spoken.

 

[Bill_Bright]

It is common with all TIMs to decrease in efficiency a few degrees over time but it should never be enough that it puts you dangerously close to thermal thresholds. That is, if you "need" those extra few degrees to keep from throttling back in speeds or reaching some other thermal protection threshold, you have other more important cooling issues to deal with, like a broken bond on your heatsink or inadequate case cooling. The point is, TIM never needs to be replaced because it is x number of years old AS LONG AS the bond between the CPU and heatsink is not broken.

Remember, there is NOT ONE, not a single solitary TIM maker, cooler maker, processor maker, motherboard maker, graphics card maker, or computer maker that says the TIM needs to be replaced on any regular or scheduled bases. Not one!

If it was needed, at least the TIM makers would recommend regular replacement as that would significantly increase their sales profits.

 

[phanbuey]

it depends on whether or not the seal breaks with liquid metal. Like if you're a laptop user and u put liquid metal between your die and the heatsink of the laptop your temps probably increased as you jostled it, and there are gaps big enough for the LM to dry and become brittle.

if you just delidded it and then resealed the IHS on the die, then your temps are probably exactly the same.

 

[Bill_Bright]

it depends on whether or not the seal breaks with liquid metal.

That's the bond I was talking about - and that's with any TIM - not just liquid metal.

for the LM to dry and become brittle.

If there are gaps that allow insulating air in, that's a problem and the old TIM definitely need to be thoroughly cleaned from the mating surfaces and a fresh new layer needs to be properly applied.

But if TIM dries, that is not a problem - contrary to what many believe. It is important to note TIM comes in a liquid form only so it can be squeezed out of the tube and spread evenly across the die. If the TIM dries out, the solids that remain behind are still there doing their jobs - filling the microscopic pits and valleys in the mating surfaces. TIM does not need to be replaced because it dries out. It only needs to be replaced if the cured bond is broken.

 

[phanbuey]

But if TIM dries, that is not a problem - contrary to what many believe. It is important to note TIM comes in a liquid form only so it can be squeezed out of the tube and spread evenly across the die. If the TIM dries out, the solids that remain behind are still there doing their jobs - filling the microscopic pits and valleys in the mating surfaces. TIM does not need to be replaced because it dries out. It only needs to be replaced if the cured bond is broken.

that's correct except in the case of liquid metal, specifically, the bond is extremely fragile, and it oxidizes when exposed to air. Additionally, gallium based LM oxidizes when used directly with copper and temps do degrade over time (my H105 was a perfect example of this).

so basically LM is NOT good when:
1) exposed to air and
2) has direct contact with a copper plate - in the case of delidding there is a layer of nickel that prevents absorption.

This is a great video of what happens and what happened to me as well:

"Liquid metal pastes react with copper heatsinks (aggravated by oxidation and higher temperatures) until the gallium in the paste is totally absorbed by the copper heatsink and you end up in thermal runaway. The damage to the copper is permanent, but if you keep reapplying the LM whenever this happens you may end up with a stable LM application. LM is much more stable when used under the heatspreader of a desktop CPU, since the heatspreader is nickle-plated and way less reactive to the LM. "

 

[Bill_Bright]

that's correct except in the case of liquid metal, specifically, the bond is extremely fragile, and it oxidizes when exposed to air.

Which would only happen if the bond is broken. So I stand by what I said. It does not need to be replaced because it is x number of years old, or if it dries AS LONG AS the bond is not broken.

 

[R-T-B]

It is common with all TIMs to decrease in efficiency a few degrees over time but it should never be enough that it puts you dangerously close to thermal thresholds. That is, if you "need" those extra few degrees to keep from throttling back in speeds or reaching some other thermal protection threshold, you have other more important cooling issues to deal with, like a broken bond on your heatsink or inadequate case cooling. The point is, TIM never needs to be replaced because it is x number of years old AS LONG AS the bond between the CPU and heatsink is not broken.

Remember, there is NOT ONE, not a single solitary TIM maker, cooler maker, processor maker, motherboard maker, graphics card maker, or computer maker that says the TIM needs to be replaced on any regular or scheduled bases. Not one!

If it was needed, at least the TIM makers would recommend regular replacement as that would significantly increase their sales profits.

Liquid metal is far from a typical TIM though. Chemically, the question is valid. It certainly reacts to a limited extent with copper, and if you put it on Aluminun heatsinks, god help you.

Which would only happen if the bond is broken.

It's a liquid compound. There isn't really a "bond" short of surface tension and a very light chemical "wetting" effect.

But if TIM dries, that is not a problem - contrary to what many believe

It is here because it becomes loose metallic crystals.

 

[Bill_Bright]

Liquid metal is far from a typical TIM though. Chemically, the question is valid. It certainly reacts to a limited extent with copper, and if you put it on Aluminun heatsinks, god help you.

What question? As I said, all TIMs degrade in efficiency over time. The question was NOT about dissimilar metals reacting with each other.

Bond? Surface tension? Regardless what you call it, it does not matter. If the computer is not bounced off the floor, or the heatsink not jerked back and forth to see if loose, the "bond" would remain intact preventing insulating air from getting in between the mating surfaces and therefore, the TIM would not need to be replaced.

It is here because it becomes loose metallic crystals.

So what? They are still there doing their jobs - taking up the space so insulating air can't get in.

 

[R-T-B]

The question was NOT about dissimilar metals reacting with each other.

It is. We're talking about liquid gallium based product mixtures. Gallinstan is probably an oldschool one you may have heard of. These modern products are similar.

So what? They are still there doing their jobs - taking up the space so insulating air can't get in.

They are loose shards of metal in a PC. I speak from experience that this is not desirable.

 

[lexluthermiester]

I have very little experience with LM TIM but I do however have a lot of experience with metallurgy. Gallium based TIM's should never be used with bare copper as the two will react and the copper will be degraded. This will cause micro-pitting and thus thermal performance degradation as a result. As is stated above, nickel plating is optimal for metals being used.

However, it is also correct that the surface bond of TIM not be broken. This is especially true with LM TIM's as the surface tension is the only bond present and is very delicate. Once applied, care needs to be taken not to disturb the application area. In the case of laptops, the fastening system employed should be enough to protect from any serious jolts. Desktop systems are a different story and can much more susceptible to jolts.

 

[Deleted member 24505]

I tried it once and found it finicky to apply, and worse to clean off.

 

[Bill_Bright]

It is. We're talking about liquid gallium based product mixtures.

It is not. Please read the title of the thread.

They are loose shards of metal in a PC. I speak from experience that this is not desirable.

Well of course loose shards of metal are not desirable. But those have absolutletly nothing to do with TIM (LM or not) that was properly applied in the first place and where the cured "bond" (or whatever you want to call it) is not broken. Beyond that, it just seems you want to drive this thread OT (which is about temps increasing over time) just for the sake of arguing. So I'm stepping out.

 

[R-T-B]

It is not. Please read the title of the thread.

"Liquid Metal TIM"

I think this is your age (of service more than physical age, mind) showing Bill, with all due respect. Liquid metal TIMs are fairly new to market. But please do read up on what they are before continuing to assume they are some type of metallic paste and not plain old Indium Gallium mixtures.

I am not driving this offtopic, rather trying to point you on-topic in a friendly manner. Applying LM TIM to Copper (or worse, Aluminum) will produce metallic crystal structures that can short a PC. It is with experience the hard way that I learned this.

 

[lexluthermiester]

It is not. Please read the title of the thread.

Liquid metal TIMs are fairly new to market.

Are there non-gallium based LM TIM's? I wasn't aware there were any..

 

[R-T-B]

Are there non-gallium based LM TIM's? I wasn't aware there were any..

There aren't. Some use Indium additives in different percentages, but they all are liquid Gallium at the core.

 

[lexluthermiester]

There aren't. Some use Indium additives in different percentages, but they all are liquid Gallium at the core.

That's what I thought. No one would be crazy enough to use an amalgam of Mercury and I can't think of any other metals that would be both good conductors of heat and liquid at room temp, regardless of alloy/amalgam composition.

 

[Bill_Bright]

It's like talking to a wall.

Liquid metal TIMs are fairly new to market.

So what? It changes nothing. READ THE TITLE TO THE THREAD! Let me repeat it just for you, R-T-B,

did your temps increase after 2 years?

It does not matter if liquid metal or not. I will say it again, all TIMs decrease in efficiency a few degrees over time. That does not mean it needs to be replaced.

And I will say this again, no TIM maker, including liquid metal TIM makers, Phobya, CoolLaboratory (which I have used) or Honey Bytes or any other say it needs to be regularly replaced!

So you are arguing for nothing.

are fairly new to market

Huh? Not hardly - fairly new would be last 2 - 3 maybe 4 years. LM TIMs have been around for at least 8 years that I know of. So fairly new to you maybe, but not me.

By the way, more years of service simply means more experience. And that experience tells me LM TIMs are not worth it - considering the prep work alone my void your CPU warranty! Yes, they can provide a few degrees improvement but so what? If you have properly set up your case cooling, the vast majority of user will never need those few degrees to prevent getting too close to thermal thresholds.

 

[R-T-B]

So you are arguing for nothing.

I'm arguing? No. I'm just stating that some advertise compatability with copper and they do technically work with it, but degrade very quickly (took mine 6 months not 2 years).

Note by degrade I mean going from "working great" to thermal shutdown. It's not just rumor either, basic chemistry can explain it.

I have no desire to argue, just to state facts. Having done that, caveat emptor.

Huh? Not hardly - fairly new would be last 2 - 3 maybe 4 years. LM TIMs have been around for at least 8 years that I know of. So fairly new to you maybe, but not me.

Meh, that's precisely what I meant. I thought maybe you were more confused than you were please disregard.

I'm not a liquid metal fan either btw, and wasn't knocking your experience at all. I do see more range than a "few degrees" but it is NOT worth the trouble for a standard build.

 

[Bill_Bright]

It's not just rumor either, basic chemistry can explain it.

No it doesn't. Improper prep and/or application sure does though.