Anyone else have problems with Arctic MX4?

[Davil]

I know that most thermal pastes are within about 1 or 2 celsius of each other, but after delidding my 4770k I'm starting to think that there is something horribly wrong with MX4. It works great for about a day or 2 and then it just deteriorates from there. I applied and reapplied it about 5 times now. It actually starts 2c lower than the AS5 I put on tonight but that has a 200 hour cure time so I'll let you guys know what the difference is in about 10 days or so. But just wondering if anyone else has had this same kind of issue with MX4.

 

[schmuckley]

Let it run for a while; mx4 is good stuff

 

[Davil]

That's the problem, I can't let it run for awhile. The temps eventually get too bad to let it run with the MX4 unless I drop my voltage.

 

[schmuckley]

Check your mounting,redo it.
You could always get some Gelid GC-extreme, too.
or Ceramique2
Gelid stuff is better.

 

[Davil]

Check your mounting,redo it.
You could always get some Gelid GC-extreme, too.
or Ceramique2
Gelid stuff is better.

Well it's hard to change out the paste without remounting so I've done that several times now. I ordered some liquid pro to put on the die since I've heard that works really well.

 

[ihog]

Are you sure you aren't putting too much on? When applying thermal paste directly to the die, you want to spread it rather than applying a dot in the center like you would on the IHS.

 

[Davil]

Yep I've actually tried the ball method, line method, and the spread method (usually do the spread for everything). And like I said it starts out great but after some time it just doesn't conduct the heat as well.

 

[Calculatron]

I recall a previous thread where someone was having a similar issue with Noctua NT-H1. Although he never found the exact problem, he claimed it was due to pump-out. However, other people have used these products and not had an issue, so I am not actually sure what to say.

Thankfully, Newegg is having a sale on Thermal compounds:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835100017&ignorebbr=1
http://www.newegg.com/Product/Product.aspx?Item=N82E16835242028&ignorebbr=1

Hopefully one of them works out, if the Arctic Silver 5 you're trying out now doesn't.

 

[Davil]

AS5 has been on there for 2 days now and I have no real change from when I put it on to now. If anything it may have gotten a little better, so I'm thinking the MX4 is either garbage on the die or it's just the tube I got.

 

[Blecter]

AS5 has been on there for 2 days now and I have no real change from when I put it on to now. If anything it may have gotten a little better, so I'm thinking the MX4 is either garbage on the die or it's just the tube I got.

Maybe a bad tube? I've got it on both my GTX 760's. One I tested before putting it on, and with the MX4 I dropped 10C under load, and the other I swapped the thermal past immediately. Peak around 71C with both cards OC'd.

 

[Calculatron]

Considering the amount of other people using it and having success, I would suspect a bad batch. It happens.

Keep us updated on the Arctic Silver 5.

 

[Dark Lord of Tech]

I have used both and prefer MX4.

 

[Brantyn Gerik]

for de-lidding applications you need to be careful which paste you use, mainly because the paste itself is subjected to SOOO much more heat when in direct contact with the cpu rather than the heatspreader. also, having something get worse over that short of time sounds more like the cooler becoming loose rather than failure of thermal paste. that being said, with mx4 having a cure time, it would make sense that immediate high heat simply just liquifies or burns it (one of the two). get a thermal paste without a cure time

 

[Davil]

Well AS5 and MX4 both have cure times. MX4 is just a few hours whereas AS5 is about 200 hours. The AS5 as of today had 0 problems both on the die and between the ihs and waterblock. I did get some liquid pro today though also and I again swapped it out. Tightening everything down in exactly the same way has given me the following results, my hottest core is within a degree of all the other ones. Under full load in prime95 the max is about 56c, and idle max is about 35c. There is a difference in ambient temperature since the last day I reported of about 3c and I really did absolutely nothing to the system other than switch the TIM. I would say that I am now a believer in liquid pro as far as putting it under the die. The difference in max temp is actually 10c-15c compared to the AS5 and MX4. I don't know whether changing out the paste above the IHS would yield similar results and maybe it's just the paste under the ihs that is the main reason it improved. I am not going to change it either because I'm quite happy with these results.

The last thing I want to stress is that I did everything as far as tightening down the IHS and the waterblock EXACTLY the same. I would be the first to say that I mounted it incorrectly with changes that drastic but I'm very meticulous in regards to mounting blocks (star pattern halfway for each one then again all around and one last check to make sure they're all tight evenly. My only conclusion is that the liquid pro works significantly better on the die.

 

[Brantyn Gerik]

yep, obviously the other two just simply arent made for the type of surface on the die vs the surface of the heat spreader