Do you really need Thermal Compound?

[kebb]

I just did the pencil-trick on my new duron 800, and now I am wondering "what if it isn't done correctly?". Is it possible to run tests of the system without the compound? Then I can always apply artic silver after I have confirmed, that the operation was succesful.

 

[Renob]

You sould always use compound I would not even try it on a overclocked cpu If you have a little tube of artic silver like I then it should last for a very large number of uses...

 

[mrchan]

do NOT run a cpu without thermal compound

 

[mikee]

Ask yourself the question - Do I really need this cpu? Then apply that answer to your original question.

 

[kebb]

I DID run a PC without Thermal Compound and it run a little hot... Although not hotter than 50c idle.

Regarding your question mikee: No I dont, but my brother does.

Thanks for the replies folks...

 

[Compellor]

You could do it if the bottom of the heatsink was very smooth and had good contact with the core. But, it's best to use some form of thermal compound to keep the chip as cool as possible. Why risk burning up the chip?

 

[monotone]

I'm using the the pink chewing-gum stuff, that was on my FOP32-1. I'm getting these temps:

Ambient: 23-24
Mobo: 26 idle, 27 load
Cpu: 42 idle, 45 load

This is on a Kt7-raid, with a Duron@900 (1.825v).

How much of a temp drop would i see if i used arctic silver?

 

[Dan]

Thermal compound is a must. There isn't any good reason not to use it.

 

[Z24]

For the short term, it's not going to hurt anything to try booting without thermal compound.

From the post, I take it that he just wants to see if it will boot? I see no problem running without thermal compound for a short period of time. Some of you make it out like if he turns the computer on without thermal compound, it will self destruct.

 

[kebb]

My intentions was as descriped in the above... and it worked out well.
(the L1-bridges was succesfully connected at first attempt)

 

[Poof]

Very lucky guy! Was it maybe Kyle at HardOCP who literally fried either a P3 or T-bird (ie., one of the earliest 1Ghz or 1.13Ghz CPUs) in seconds when the HS didn't make good contact with the core? They're still talking about it to this day...

Anybody recall? (I'm lazy and don't feel like going over there to find the review... )

 

[Mikewarrior2]

It was a 1ghz t-bird, with a PEP66(the feet were too tall for cpu core) and it melted in 8Secs.


Mike

 

[Bigdude]

I had a Cyrix 166 a few years back that I had a heatsink on it, but forgot to plug in the fan, it lasted about 3mins. Vendor gave me a new one no questions asked.

 

[Turbopit]

Why would you not want to use it? I see no reason? it's there to take up any air gaps.

 

[bonkers325]

yes u do if u want good contact between cpu and heatsink

 

[JmsAndrsn]

It's almost like driving your car without motor oil. I wouldn't recommend it.

 

[ninjazed]

What about the conductive patch on the bottom of many heat sinks? Isn't that the same as thermal compound if a good contact is made with the CPU? I just installed a celeron 633 in a friends box and the center core is raised a little from the rest of the chip anyway. There is direct contact between the core and the heat sink patch. Whaddayathinks about that?

 

[Mikewarrior2]

Thermal pads vary greatly.

On Approved Socket-A heatsinks, thermal pads are in hte form of the Chometrics T725(or something like that) PCTC pad. They perform fairly well.

ON other heatsinks, hte pad is generally junk.


Mike