matze79 wrote:

Superglue (Acrylacrylate) ....

I believe it's cyanocrylate.

KT7AGuy wrote:

... it's a real crappy solution.

I know lots of mobos that can attest to the contrary, but to each his own. 😀

TELVM wrote:

I know lots of mobos that can attest to the contrary, but to each his own. 😀

I've never meet a mobo that could write or speak.

PCBONEZ wrote:

Personally I'm opposed to supporting companies that lie to their customers.
After learning the truth about AS5 I won't be supporting Arctic Silver by buying any of their products.
.

That analysis you posted of AS5 is interesting. The stuff tends to perform quite well in all testing I've seen elsewhere. It actually outperforms the Shin Etsu X23 (here) after curing! Maybe there is more to it. I can't tell if the study took the curing time into consideration. They also tested with a gap of at least 25 microns.

Regardless, you are right that it is overpriced. Lots of better values out there.

saturn wrote:

matze79 wrote:

Isnt Arctic Silver 5 pretty outdated ?

it sure is. NT-H1, mx-4 and coollaboratory liquid are the only tim worth buying if you ask me. other then that, just use what comes with you heatsink.

I had some "Cooler Master" branded paste that came with a heatsink. After using it a few times, even on some short lived builds, I kept finding it to later have the qualities of chalk dust. It's so worthless that I don't think I'll be using it anymore for anything, so I should probably just throw it away.

Unless you go crazy, typical thermal paste of some decent brand is cheap. There's an up front cost of a few bucks and then the tube lasts most normal people for years or a lifetime. For people around here maybe not quite as long, but still, the cost per build is totally negligible.

I've gotten used to using AS Ceramique because it just works fine and it's widely available and cheap. I'm sure there's many other decent brand pastes that are just as good. I'm skeptical of what value the expensive pastes really offer, I just don't want something that turns into chalk.
I don't know who makes/made Radio Shack paste, but I was happy with that too, I had a tube of it for years that I paid 99 cents for.

swaaye wrote:

PCBONEZ wrote:

That analysis you posted of AS5 is interesting. The stuff tends to perform quite well in all testing I've seen elsewhere. It actually outperforms the Shin Etsu X23 (here) after curing! Maybe there is more to it. I can't tell if the study took the curing time into consideration. They also tested with a gap of at least 25 microns.

There is definitely more to it, but among all the A / A+ / A- TIMs in that table, there is no statistically significant difference. So while it is probably fair to determine a list of good solutions, you cannot really say any one of them outperforms any other.

There may be differences between them in non-perfect conditions, or in terms of longevity (Does it dry out and performs worse? Does it need to be replaced every once in a while?) These are not reflected in that study.

PCBONEZ wrote:

I've never meet a mobo that could write or speak.

You need to go out more. 🤣

/\ That mobo desperately needs a true speaker 🤣

dr_st wrote:

swaaye wrote:

PCBONEZ wrote:

That analysis you posted of AS5 is interesting. The stuff tends to perform quite well in all testing I've seen elsewhere. It actually outperforms the Shin Etsu X23 (here) after curing! Maybe there is more to it. I can't tell if the study took the curing time into consideration. They also tested with a gap of at least 25 microns.

There is definitely more to it, but among all the A / A+ / A- TIMs in that table, there is no statistically significant difference. So while it is probably fair to determine a list of good solutions, you cannot really say any one of them outperforms any other.

There may be differences between them in non-perfect conditions, or in terms of longevity (Does it dry out and performs worse? Does it need to be replaced every once in a while?) These are not reflected in that study.

Well, since many of these tests are done when the TIM has been applied very recently, perhaps theres a difference in how long a TIM stays effective?

If this is the case, I'd very much prefer a TIM that may be slightly less effective at first, but has more of a tendency to stay that way for a longer time. With some coolers I really dislike having to reapply the TIM because I have to dismantle half the rig just for that 🤣!

Tetrium wrote:

There may be differences between them in non-perfect conditions, or in terms of longevity (Does it dry out and performs worse? Does it need to be replaced every once in a while?) These are not reflected in that study.

This reminds me of one hilarious TIM review.
http://www.dansdata.com/goop.htm

The moral of this story is (as you say) plain zinc oxide or silicon grease will work just fine. As for longevity, I'm not convinced of the utility. Conventional grease is proven to work for many years of service in all manner of high heat, hard duty electronics systems. In fact, I'm somewhat surprised the audiophile crowd has not latched on to exotic thermal compounds for improving system performance as they're normally drawn to dubious claims like flies to bullshit.

If you want to get exotic, try something from the condiments aisle.

gdjacobs wrote:

Tetrium wrote:

There may be differences between them in non-perfect conditions, or in terms of longevity (Does it dry out and performs worse? Does it need to be replaced every once in a while?) These are not reflected in that study.

This reminds me of one hilarious TIM review.
http://www.dansdata.com/goop.htm

Thats nothing. http://forums.overclockersclub.com/topic/1644 … u-in-my-peanut/\
http://forums.overclockersclub.com/topic/1644 … anut/?p=1680970

dr_st wrote:

swaaye wrote:

PCBONEZ wrote:

That analysis you posted of AS5 is interesting. The stuff tends to perform quite well in all testing I've seen elsewhere. It actually outperforms the Shin Etsu X23 (here) after curing! Maybe there is more to it. I can't tell if the study took the curing time into consideration. They also tested with a gap of at least 25 microns.

There is definitely more to it, but among all the A / A+ / A- TIMs in that table, there is no statistically significant difference. So while it is probably fair to determine a list of good solutions, you cannot really say any one of them outperforms any other.

There may be differences between them in non-perfect conditions, or in terms of longevity (Does it dry out and performs worse? Does it need to be replaced every once in a while?) These are not reflected in that study.

Yeah you are right. I like that roundup because it shows how minimal the differences really are among most of the TIM options available. I also liked their testing methodology. They took curing time, surface aspects and application techniques into consideration.

gdjacobs wrote:

In fact, I'm somewhat surprised the audiophile crowd has not latched on to exotic thermal compounds for improving system performance as they're normally drawn to dubious claims like flies to bullshit.

A mystery indeed. I see opportunity. 🤣

swaaye wrote:

gdjacobs wrote:

In fact, I'm somewhat surprised the audiophile crowd has not latched on to exotic thermal compounds for improving system performance as they're normally drawn to dubious claims like flies to bullshit.

A mystery indeed. I see opportunity. 🤣

🤣 wow! But there is a problem; anyone can see how good the tim is by looking at the temps of the cpu. Unlike the fancy cables which gives a placebo affect. One would have to come up with some fancy wording to trick them into thinking that the cpu is cooler then what the sensors say. Or play on their foolishness.

Actually the study swaaye linked to contradicts is self as to if they followed curing time instructions.
In one place it says they followed them but in another it says
-"Since each test consumes an average 90-minutes from setup to tear-down, "-
You can't do a 48 hour or 200 hour cure time in 90 minutes.
Further, with 80 samples if only one had a 200 hour cure time and all the rest were one hour the average would be at least 3.5 hours.

While I also liked their test methodology over all it has a fatal flaw.
They used the same thermocouples (mobo/CPU) for 16 months to 2 years (depending on where you read) without ever checking them for consistency.
Thermocouples are especially prone to "drifting" which means their properties change over time.
Thus there is zero confidence that what the TC was outputting for (say) 80 degrees at the beginning was the same as it was outputting for 80 degrees at the end.
It would have been very easy to check and they simply didn't do it.

That there were 80 products and only about 5 degrees difference from the absolute best to the absolute worst tells me there was something wrong with their methodology, practice, math or something.
I don't know what was wrong but an outcome like that is simply not likely. Possible - yes, likely - no.
IMHO the 'spread' should be a lot wider.

-

In response to comments about the study I linked to.
You have to bare in mind that what you see online is just a summary. A teaser. They want you to go to the link provided and buy the full study.
They INTENTIONALLY aren't telling you everything.
If you look it DOES say one of the things were looking at was aging. (Aging results just don't happen to be in the teaser.)
Clearly, if they are age testing the stuff the curing requirements were met.

They also say their tests were done IAW ASTM D5470. - ASTM D5470 is an Industry Standard test method.
Part of my job as a QA Inspector (for submarines - the whole submarine, not just electronics/nuclear) was to write (create/design) tests based on things like ASTM, Mil-Spec, Mil-STD, ANSI and a dozen others. I was in those books all the time.
For something like this I guarantee you an ASTM specifies to follow the manufacturers instructions.

As a Laboratory would be using calibrated scientific grade RTDs and checking the calibration repeatedly, drift would not be an issue for them.
.

RTDs are resistive whereas thermocouples generate a voltage output as a function of temperature differential, but you're correct in that they are far superior (especially 4 lead RTDs) within their temperature range. Thermocouples are sometimes the only option available at extreme temperatures.

I've worked from ASTM manuals as well. Depending on the test in question, they can vary quite a lot in terms of thoroughness, but I suspect D5470 is pretty good.

gdjacobs wrote:

RTDs are resistive whereas thermocouples generate a voltage output as a function of temperature differential, but you're correct in that they are far superior (especially 4 lead RTDs) within their temperature range. Thermocouples are sometimes the only option available at extreme temperatures.

I've worked from ASTM manuals as well. Depending on the test in question, they can vary quite a lot in terms of thoroughness, but I suspect D5470 is pretty good.

Yes. You are right. I should edit a little so I don't confuse anyone.

I don't recall the spec or standard but we had this damned book about O-rings. Over 900 pages, little tiny print and ALL about O-rings. Everybody hated that thing.

However, as NASA and Thiokol learned, God help you if they fail.

gdjacobs wrote:

However, as NASA and Thiokol learned, God help you if they fail.

Actually after NASA had their little problem they adopted our QA program.
The higher-ups were all proud of that but it didn't matter much to me.

PCBONEZ wrote:

... Part of my job as a QA Inspector (for submarines - the whole submarine, not just electronics/nuclear) ...

Wow, that's cool . Boomers or fast attacks?

How well you apply the thermal paste will probably have far more bearing on the outcome than the content of the thermal paste.

No thermal paste is as good at transferring heat as "actual" metal-to-metal contact.

Emphasis on "actual". Thermal paste is only there to fill the microscopic air gaps caused by the CPU and HSF's surfaces not being perfectly flat. Where those gaps exist, heat must travel through air. While thermal paste isn't as good as metal-to-metal contact, it's shitloads better than air at transferring heat.

Also: http://www.dansdata.com/goop.htm tl;dr use toothpaste or Vegemite. 🤣

TELVM wrote:

Wow, that's cool http://smileys.on-my-web.com/repository/Respect/respect-048.gif . Boomers or fast attacks? […]
Show full quote

PCBONEZ wrote:

... Part of my job as a QA Inspector (for submarines - the whole submarine, not just electronics/nuclear) ...

Wow, that's cool . Boomers or fast attacks?

I did both but the facility I was at mostly did Boomers so I saw a lot more of those.