TPU's Nostalgic Hardware Club

[debs3759]

After the initial surprise of this being such an easy fix, I opened up the BIOS and, holy-moly, this thing had an Athlon XP 3000+ inside! The CPU used to be the king of the market in the early 2000s (price at launch was 536 $). One thing that alarmed my senses was the CPU temperature. It was idling from 55-57 °C. I shut the whole PC off and went ahead to change the thermal compounds on both the CPU and GPU coolers. I used my old trustworthy ArcticCooling MX-2 for the job.

(CPU thermal paste application)
View attachment 80448

(GPU thermal paste application)
View attachment 80449
After re-applying the thermal paste, putting back the CPU and GPU coolers, dusting the PC, wiping it a little bit, adding a fan infront of the case and checking if everything was in it's place it was time to try and put an OS onto it. This is what I was dealing with at the time:

Is that before or after you cleaned and reapplied the TIM? If after, you might try just putting an amount equivalent to a grain of rice or a small X in the middle of the die. Too much TIM is almost as bad as too little.

MX-2 is my go-to paste as well

 

[Mr.Scott]

Bare die CPU's should be spread, paper thin over the entire die to avoid hot spots. You can't trust a dot or x to cover the entire die on compression.

 

[Thimblewad]

Thats'a great find you ran across - Thanks for sharing it.
The XP-3200 200MHz chip was the actual king back in the day related to desktop CPU's and I have two of them here, both do well and one happens to be a pre-superlock dated chip. However in reality the XP-M's are the real stars of Socket A but unfortunately not all Socket A boards support them. Some will, some will also but not properly and even some won't run them at all, depends on the board.

Now - You found a great setup and the price was right, I'd try a few rounds of 3D Mark 01, maybe a round or two of PC Mark 04 along with the usual stuff and see what you can make it do.

Well, I actually got this PC for free, the RAM was just waiting there for years, so I actually didn't put anything into it that wasn't already in my room somewhere. And I sleep in here, haha

Is that before or after you cleaned and reapplied the TIM? If after, you might try just putting an amount equivalent to a grain of rice or a small X in the middle of the die. Too much TIM is almost as bad as too little.

MX-2 is my go-to paste as well

That is after the TIM is re-applied, but the layer is actually very thin (if I would try to move the paste, my finger would clean off a certain part of it) and I rather just align it evenly with my fingers. I don't know, that's my way of doing it, hahaha.

As for the MX-2: I saw potential in it when I bought my CPU cooler with the paste pre-applied, I was like, wow. And from there on I always had great results when I applied it for the first time and it has never failed me yet. The temperatures are a lot lower comparing to the manufacturer TIM, my findings.

 

[Trekkie4]

For those of you who followed all my projects over the last couple of months (and years), here's a blast from the past!

Whatever you do, please try not to apply TIM like this... (yup, still the worst one I've ever encountered!)

 

[dorsetknob]

3 Different Tims ? and some fur @peche going to be visiting here soon asking for temps spreading method ect

 

[DR4G00N]

My Cele 1GHz & Biostar M6VLQ.


It's now running a PIII-S 1.4GHz @ 1.6GHz 1.7V with a Hyper 212+ lol.

 

[Trekkie4]

3 Different Tims ? and some fur @peche going to be visiting here soon asking for temps spreading method ect

Lol, most likely yeah! The guy who did this probably ran out of TIM so he had to improvise any way he could! I'm probably being too harsh on him, especially since I was given the entire system for free... I'm sure the guy had best intentions but oh man!! It took me ages and half a bottle of WD-40 to get all this gunk off the CPU! Like I said earlier, still the worst case of TIM I've EVER seen before in my life!

 

[GreiverBlade]

For those of you who followed all my projects over the last couple of months (and years), here's a blast from the past!

Whatever you do, please try not to apply TIM like this... (yup, still the worst one I've ever encountered!)

and i thought my 1st self assembled rig TIM application was a mess ... (K6-2 450) or my second rig with a A-XP ...

 

[Trekkie4]

Here are all 5 pics from the original post (#3569, page 143) CPU model was (and still is) AMD Athlon 1000, A1000AMT3C

 

[Mr.Scott]

That's outstanding.

 

[stinger608]

Bare die CPU's should be spread, paper thin over the entire die to avoid hot spots. You can't trust a dot or x to cover the entire die on compression.

I totally agree. What I always did was take a sandwich baggie and turn it inside out. I'd then put a small dab on the cooler smear it using the baggie on the cooler itself. Always seemed to work well.

 

[peche]

3 Different Tims ? and some fur @peche going to be visiting here soon asking for temps spreading method ect

and also redirecting people to a survey post! thanks for the mention !
Lol!

Regards,

Lol, most likely yeah! The guy who did this probably ran out of TIM so he had to improvise any way he could! I'm probably being too harsh on him, especially since I was given the entire system for free... I'm sure the guy had best intentions but oh man!! It took me ages and half a bottle of WD-40 to get all this gunk off the CPU! Like I said earlier, still the worst case of TIM I've EVER seen before in my life!

tell us moar about the paste and thermals ! lolz!

I totally agree. What I always did was take a sandwich baggie and turn it inside out. I'd then put a small dab on the cooler smear it using the baggie on the cooler itself. Always seemed to work well.

+1 here, great advise, also great pic! see you really like it!

Regards,

 

[Trekkie4]

@peche Not sure I understood your question, you mean temp results with that amount of TIM all over the CPU? It's been a long time since then, but from what I remember system DID boot up & worked within acceptable temps, otherwise it would most likely overheat and shut down. So much for various TIM spreading techniques & other nonsense, even with all this mess it STILL worked OK!

But yeah, as you can tell from the pics, I removed all the TIM from the CPU using WD-40 & applied fresh coat over the CPU die. I also swapped the original AGP card from FX5700LE to Evil Kyro & did few more tweaks since then, but otherwise it's still operational & mostly "original". The guy who owned this system built it himself back in a day (certainly explains a thing or two), but due to moving into smaller house he had to either scrap it or give it to someone who will love it & hopefully restore it to former glory Which I hopefully did!

 

[peche]

@peche Not sure I understood your question, you mean temp results with that amount of TIM all over the CPU? It's been a long time since then, but from what I remember system DID boot up & worked within acceptable temps, otherwise it would most likely overheat and shut down. So much for various TIM spreading techniques & other nonsense, even with all this mess it STILL worked OK!

But yeah, as you can tell from the pics, I removed all the TIM from the CPU using WD-40 & applied fresh coat over the CPU die. I also swapped the original AGP card from FX5700LE to Evil Kyro & did few more tweaks since then, but otherwise it's still operational & mostly "original". The guy who owned this system built it himself back in a day (certainly explains a thing or two), but due to moving into smaller house he had to either scrap it or give it to someone who will love it & hopefully restore it to former glory Which I hopefully did!

forget it, it was a completely joke about temps, @dorsetknob always make fun of my weird thermal fetish!

Regards,

 

[Trekkie4]

Oops, sorry! (apparently I completely missed the point )

 

[Thimblewad]

@Trekkie4 Damn, and I thought my first attempt was bad, good job on cleaning it, must have taken a bucket-load of time!

 

[Trekkie4]

He-he, a lesson for all of us to be learned... No matter how bad something might seem, it can ALWAYS be (and get) worse!

 

[buzzi]

did that guy directly dip the cpu in TIM? it looks like rescued from mug

 

[Trekkie4]

did that guy directly dip the cpu in TIM? it looks like rescued from mug

IDK about that, but he probably assumed how adding extra TIM would result with extra cooling efficiency Good thing he didn't smear all this stuff across the heatsink & under the fan - THAT would have been a nightmare to clean!

Oh well, still better than THIS guy, in the pic below!!

Edit
Or even better, this one here!

 

[buzzi]

that's the reason why TIM manufacturers decided to produce electrically non-conductive ones

 

[GreiverBlade]

IDK about that, but he probably assumed how adding extra TIM would result with extra cooling efficiency Good thing he didn't smear all this stuff across the heatsink & under the fan - THAT would have been a nightmare to clean!

Oh well, still better than THIS guy, in the pic below!!

Edit
Or even better, this one here!

actually the second one make almost some sense if it was for a TEC or a phase changer cooling ... heck even the 1st one, don't judge before knowing the full story (unless you know the full story and want to share )

http://forums.pureoverclock.com/work-logs/5810-thermoelectric-cooling-icekool-8th-april-2.html
Die-electric grease applied to the CPU socket. (di-electric mean non conductive in French should be the same in English actually the op did a slight typo : Die=/=di)

notice the 2 pics on the left, it's from a french computer review on the Asetek Vapochill install, well a ... nostalg33k paper ...


notice the socket:


(it's not even that old ... this practice well yes it is ... but it's still used ) actually a TEC/phase changer cooler can go sub-zero in idle that's why you need TIM in the socket to avoid icing (under the cake)

this is what you don't want :
"Originally Posted by [ShowMe!]

That sucks!

Try to wash the components, dry them out really well and see if you can get them back up running.

When I built my chill-box air conditioner chiller. The inside of the chillbox never came up above freezing (32f) even during the off period at night, this prevented any moisture to build up.

I did have leaks in my chillbox due to poor construction, because it was for a test only, however I noticed that the small amount of frost that formed concentrated it self on the radiator that was cooling the inside air, and the rest of the components remained dry and clean.

I think the problem with insulation is that its hard to be 100% no matter what you do, the thermal cycling of the components will create low and high temperature areas, that will lower and raise the pressure around them, and I think it acts like a slow air pump that pulls humid air and traps it inside, and with each cycle more and more moisture accumulates.

I think people put some type of di-electric grease right in to the socket, you have to read up on this because I never done it before.





What's the gloppy stuff? It looks like my CPU after I embedded it in silicone, except yours isn't plugged into the socket(!)



If it's dielectric, wouldn't that cause insulation between the pin and the socket?



I sealed my CPU and waterblock airtight under the rubber, figuring that the only water condensing would be what's there already. I don;t know if that's valid. But it doesn;t take much silicone to prevent ALL external condensation. It shapes itself to the bottom of the waterblock and whatever lives under the socket. It also peels off when you want. I'd use that instead of eraser.



-faith"
from http://www.overclock.net/t/1526237/the-downside-of-using-a-phase-change-and-or-extreme-cooling/20

 

[Trekkie4]

Wow, never heard of this wizardry before And I thought how water cooling (and peltier elements) are extreme solutions to keeping your PC cool & quiet

Of course, this is obviously not a common practice and wouldn't work for just about anyone. Especially me, I already have a lot of problems with humidity as it is...

Also, call me old-fashioned, but I would never trust anything (including TIM) to get between the CPU and mobo socket. Over the time, this thing will gather dust & contaminate itself, to the point where it'll either insulate pins from touching the base OR become conductive & short all the pins together... I wouldn't mind testing it on an old motherboard, such as S462 or 370, but risking a brand new motherboard with Core i7 for example would be out of the question.

Furthermore, let me ask you a question... Suppose you DO make it work & have it set up for a while but eventually decide to swap it back for the original specs, or even sell the mobo & CPU - how the heck would you clean CPU pins without damaging them?!

 

[DR4G00N]

Haha, let me tell you that cleaning die-electric grease out of an lga socket is not very fun... but it can be done.

My Evga e762 is covered in the stuff because I planned to use it with dry ice last year (ended up using my X58A-OC instead) nothing bad has happened yet from all the dust sticking to it. It's getting pretty nasty though so I'll probably clean it off soon lol.

 

[Trekkie4]

Haha, let me tell you that cleaning die-electric grease out of an lga socket is not very fun... but it can be done.

My Evga e762 is covered in the stuff because I planned to use it with dry ice last year (ended up using my X58A-OC instead) nothing bad has happened yet from all the dust sticking to it. It's getting pretty nasty though so I'll probably clean it off soon lol.

Right, but you're missing a point... Suppose dust (or other particles) find their way between the CPU and mobo contacts, it would insulate the contact(s) from touching one another, which translates to non-working system. Especially if (and when) you decide to upgrade the CPU or change it for whatever the reason. All the gunk would fall into the pin holes (assuming you're using AMD platform, which uses old-fashioned design and not the inverted one like Intel) Sooo yeah, not worth the risk, but that's just my, personal opinion.

There is also a huge risk of shorting something if and when metal scrapings get trapped inside. And you can find metal scrapings everywhere, either from stripping a screw threads, mounting the hard drive into the 3.5" bay, etc. IDK, it just seems like a really REALLY bad idea *shrug*

 

[Mr.Scott]

All of my benchers have dielectric in the sockets. It doesn't hurt anything. Some of them have been that way for the better part of 10 years.