Thuban delidded

[chew*]

So I was fooling around in this giga contest with what apparently was a problematic board.

Usually I would just stop when a board starts acting up but for the sake of competition I plodded on through all the issues the board had although i did break down 10 times and dry it out just to get 2 results in wprime..........eventually as I expected the board died, unfortunately it took the cpu with it it seems.

Bad for me but hey even dead cpu's have a purpose right? Lets see if thuban can safely be delidded.

Included deneb for comparison. As we can see the die is larger and it appears they are now using some type of glue around surrounding chips and running them so close to the IHS glue point that cutting through it without nicking one is near impossible.






I think it's safe to say for users that do delid chips that they skip delidding thuban and just lap it really really thin.

[Particle]

Very cool photos, chew. I loves me some nekked Si. I'm surprised to see the die area isn't really all too much larger. It really shows off just how tiny AMD's K10 cores really are.

[crazydiamond]

cool pics, thanks chew

[FatAlbert]

thuban without IHS looks a little bit ugly :P

[Manicdan]

it looks like its a little wider and a little tall, can you get out the ruler please?

[chew*]

it looks like its a little wider and a little tall, can you get out the ruler please?

Thuban





Deneb

[demonkevy666]

why is it so SHINNY @_@ ?

the deneb looks dull. lol

[chew*]

why is it so SHINNY @_@ ?

the deneb looks dull. lol

Becasue i cleaned it with a fresh razor really really good.

The TWKR "deneb" in the lowest picture if you look close has a cracked core....I didn't clean deneb, instead i sanded it as i had a feeling i cracked core( harder to find crack when shiny so i used rough sandpaper ), found the crack and I knew why

[Ket]

Sooo.... Shi...ny.....

[FlanK3r]

only about 2mm in lengt diference? Nice photos too.

[demonkevy666]

Becasue i cleaned it with a fresh razor really really good.

The TWKR "deneb" in the lowest picture if you look close has a cracked core....I didn't clean deneb, instead i sanded it as i had a feeling i cracked core( harder to find crack when shiny so i used rough sandpaper ), found the crack and I knew why

nice job.

I was expecting it to be just a bit longer, not wider too.

since also it's wider, it must help dissipate heat better with more surface area.

[Royalk]

yaaaaaaaaaaaaa....I like this. Still alive?

[BeepBeep2]

yaaaaaaaaaaaaa....I like this. Still alive?

He said because of the glue on the little components around the core you cant cut through safely...I think the glue sticks to the IHS and pulls the component out.?

[chew*]

He said because of the glue on the little components around the core you cant cut through safely...I think the glue sticks to the IHS and pulls the component out.?

No the components are so close to the rubber they use to hold IHS on that to cut through it all the way you will almost always guranteed knick one of those SMD's.

Knicking one would be very very bad on a working cpu.

[BeepBeep2]

Can you measure the diameter of the rubber? ...would it be possible to cut slightly less than that diameter then rip the IHS off?

[chew*]

Can you measure the diameter of the rubber? ...would it be possible to cut slightly less than that diameter then rip the IHS off?

Well the problem lies in the fact that the IHS is soldered to core. Any attempt to "pop" it while still soldered will crack core.

In order to get it to "pop" the rubber needs to be cut all the way through other wise solder melts and then just rebonds.

[BeepBeep2]

Well the problem lies in the fact that the IHS is soldered to core. Any attempt to "pop" it while still soldered will crack core.

In order to get it to "pop" the rubber needs to be cut all the way through other wise solder melts and then just rebonds.

Ahh, I see...that makes more sense.

Why do they make them so hard to get off?

[chew*]

Ahh, I see...that makes more sense.

Why do they make them so hard to get off?

cores are fragile and easily cracked?

[BeepBeep2]

cores are fragile and easily cracked?

I mean, why do they solder the IHS to the core...If somebody is going to cut the rubber to pull off the IHS then what's the point of the solder? They end up killing their CPU.

[chew*]

I mean, why do they solder the IHS to the core...If somebody is going to cut the rubber to pull off the IHS then what's the point of the solder? They end up killing their CPU.

Paste dries out over time? solder doesn't and provides a better thermal interface.

[BeepBeep2]

Paste dries out over time? solder doesn't and provides a better thermal interface.

I'm full of curiosity tonight... (sorry I'm asking 20 questions...)
Does paste still dry out if the chip is sealed? (Phase change material even?)

I agree about the solder providing a better thermal inferface.

[chew*]

I'm full of curiosity tonight... (sorry I'm asking 20 questions...)
Does paste still dry out if the chip is sealed? (Phase change material even?)

I agree about the solder providing a better thermal inferface.

yep it dries, i pulled lid off an old athlon 939 chip recently, it gets preety powdery.

[Daveburt714]

Thanks for the pics chew*...

Looks like their pretty close to the limit of cores that will fit on an AM3 package with K10.

[BeepBeep2]

I could maybe see eight squeezed in there. The die could definately get a little wider...Or maybe a small boost in L2 or L3. ? maybe I speak nonsense because there is no spoon. Who knows?

[richierich]

yep thanks for the pics and info, the Thuban will be missed lol