Thuban delidded

Printable View

07-23-2010, 12:42 PM
chew*

Thuban delidded

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.

http://chew.ln2cooling.com/890FXA-UD...lidded%201.jpg

http://chew.ln2cooling.com/890FXA-UD...lidded%202.jpg

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.
07-23-2010, 01:05 PM
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.
07-23-2010, 01:07 PM
crazydiamond

cool pics, thanks chew
07-23-2010, 01:14 PM
FatAlbert

thuban without IHS looks a little bit ugly :P
07-23-2010, 01:18 PM
Manicdan

it looks like its a little wider and a little tall, can you get out the ruler please?
07-23-2010, 01:36 PM
chew*

Quote:

Originally Posted by Manicdan

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

Thuban

http://chew.ln2cooling.com/thuban/th...lidded%203.jpg

http://chew.ln2cooling.com/thuban/th...lidded%204.jpg

Deneb

http://chew.ln2cooling.com/thuban/th...lidded%205.jpg

http://chew.ln2cooling.com/thuban/th...lidded%206.jpg
07-23-2010, 01:49 PM
demonkevy666

why is it so SHINNY @_@ ?

the deneb looks dull. lol
07-23-2010, 02:01 PM
chew*

Quote:

Originally Posted by demonkevy666

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 ;)
07-23-2010, 02:27 PM
Ket

Sooo.... Shi...ny..... :slobber:
07-23-2010, 02:38 PM
FlanK3r

only about 2mm in lengt diference? Nice photos too.
07-23-2010, 02:49 PM
demonkevy666

Quote:

Originally Posted by chew*

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.
07-23-2010, 05:15 PM
Royalk

:shocked:yaaaaaaaaaaaaa....I like this. Still alive?
07-23-2010, 05:25 PM
BeepBeep2

Quote:

Originally Posted by Royalk

:shocked: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.? :shrug:
07-23-2010, 05:48 PM
chew*

Quote:

Originally Posted by BeepBeep2

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.? :shrug:

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.
07-23-2010, 05:53 PM
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?
07-23-2010, 05:56 PM
chew*

Quote:

Originally Posted by 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?

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.
07-23-2010, 06:49 PM
BeepBeep2

Quote:

Originally Posted by chew*

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? :confused:
07-23-2010, 06:51 PM
chew*

Quote:

Originally Posted by BeepBeep2

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

Why do they make them so hard to get off? :confused:

cores are fragile and easily cracked?
07-23-2010, 07:06 PM
BeepBeep2

Quote:

Originally Posted by chew*

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. :confused:
07-23-2010, 07:09 PM
chew*

Quote:

Originally Posted by BeepBeep2

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. :confused:

Paste dries out over time? solder doesn't and provides a better thermal interface.
07-23-2010, 07:15 PM
BeepBeep2

Quote:

Originally Posted by chew*

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

I'm full of curiosity tonight...:p: (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.
07-23-2010, 08:10 PM
chew*

Quote:

Originally Posted by BeepBeep2

I'm full of curiosity tonight...:p: (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.
07-23-2010, 08:38 PM
Daveburt714

Thanks for the pics chew*... :D

Looks like their pretty close to the limit of cores that will fit on an AM3 package with K10.
07-23-2010, 09:18 PM
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? :D
07-24-2010, 01:46 AM
richierich

yep thanks for the pics and info, the Thuban will be missed lol
07-24-2010, 03:57 AM
Ket

Quote:

Originally Posted by 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? :D

Matrix fan :p:
07-24-2010, 11:04 AM
apt403

Anyone ever try to use that freeze spray to freeze the glue around the IHS? I've managed to shatter rubbing insulation with that stuff before, might aid in the removal of the heatspreader.
07-24-2010, 12:27 PM
BeepBeep2

Quote:

Originally Posted by apt403

Anyone ever try to use that freeze spray to freeze the glue around the IHS? I've managed to shatter rubbing insulation with that stuff before, might aid in the removal of the heatspreader.

Or dip the chip in LN2? :)

I had to :ROTF:
07-24-2010, 01:01 PM
chew*

Quote:

Originally Posted by apt403

Anyone ever try to use that freeze spray to freeze the glue around the IHS? I've managed to shatter rubbing insulation with that stuff before, might aid in the removal of the heatspreader.

Quote:

Originally Posted by BeepBeep2

Or dip the chip in LN2? :)

I had to :ROTF:

So the experiments continue but i already know the answer to this one, theres a reason that other deneb core cracked ;)

Ln2 dipping test take 10.

Right after defrosting

http://chew.ln2cooling.com/890FXA-UD...0delid%207.jpg

Wiped mosture off.

http://chew.ln2cooling.com/890FXA-UD...0delid%208.jpg

http://chew.ln2cooling.com/890FXA-UD...0delid%209.jpg

http://chew.ln2cooling.com/qdig-file...delid%2010.jpg
07-24-2010, 01:05 PM
FatAlbert

looks very bad :/
07-24-2010, 01:08 PM
chew*

Quote:

Originally Posted by FatAlbert

looks very bad :/

yah looks like cores preety fragile. I wouldn't reccomend LN2 dipping to remove IHS, core will crack in 8 places :rofl::ROTF:
07-24-2010, 01:12 PM
FatAlbert

:rofl:
07-24-2010, 01:35 PM
crazydiamond

hmm wonder what it looks like under there ;)
07-24-2010, 02:10 PM
chew*

Quote:

Originally Posted by crazydiamond

hmm wonder what it looks like under there ;)

ahh shall i tear apart silicon now? Why not.
07-24-2010, 03:14 PM
knopflerbruce

Dead chips aren't useless. I found out F3 Windsors were soldered that way:D

Are Regor chips soldered too, btw?
07-24-2010, 03:28 PM
chew*

Quote:

Originally Posted by knopflerbruce

Dead chips aren't useless. I found out F3 Windsors were soldered that way:D

Are Regor chips soldered too, btw?

I believe since all chips are native denebs for he most part now the answer is yes.

So on to the photo shoot. I might actually break out my macro lens if i can find it.

Lets see what i can pull off without the macro lens with my point and shoot however.

Dialing in zoom white balance is not correct yet.

http://chew.ln2cooling.com/thuban/th...delid%2012.jpg

Lets try a smaller piece.

http://chew.ln2cooling.com/thuban/th...delid%2011.jpg

even smaller

http://chew.ln2cooling.com/thuban/th...delid%2013.jpg

Tad to much white balance.

http://chew.ln2cooling.com/thuban/th...delid%2014.jpg

There's the money shot :)

http://chew.ln2cooling.com/thuban/good%20shot%20.jpg
07-24-2010, 03:33 PM
Deluded

Ah, those just came out... you heartless murderer, killing babies like that....! They're only kids... KIDS! http://forums.extremeoverclocking.co...s/azcrying.gif

So much eager life lost.... :mad:

:p:
07-24-2010, 03:35 PM
Deluded

Holy :banana::banana::banana::banana:. that is scary. It feels like I'm looking at something that I'm not supposed to.
07-24-2010, 04:15 PM
crazydiamond

that last shot is just awesome! thanks!
07-24-2010, 04:23 PM
thijs

awesome chew, last shot is really epic :) looks cool. a worthy sacrefice for sure!
07-24-2010, 04:38 PM
cdawall

Cool :banana::banana::banana::banana: chew* so what about heating and rapidly cooling?
07-24-2010, 04:46 PM
chew*

Quote:

Originally Posted by cdawall

Cool :banana::banana::banana::banana: chew* so what about heating and rapidly cooling?

I imagine it would make it worse.

If anything you would want to slow down the cooling/warming process so everything has time to adjust.
07-24-2010, 04:49 PM
Chumbucket843

Quote:

Originally Posted by chew*

cores are fragile and easily cracked?

YES!

it's basically glass. i can teach you how to do die deprocessing if you would like. ive done it on some chips.

edit: too late :(

you can further process the metal layers.
07-24-2010, 04:57 PM
chew*

Quote:

Originally Posted by Chumbucket843

YES!

it's basically glass. i can teach you how to do die deprocessing if you would like. ive done it on some chips.

edit: too late :(

you can further process the metal layers.

I have another dead deneb here that it can be tried on but from my experience they have a lot harder core.

I just pounded one into oblivion earlier.

I couldn't even get as close to a clean die shot as i was able to with thuban.
07-24-2010, 05:14 PM
cdawall

Quote:

Originally Posted by chew*

I imagine it would make it worse.

If anything you would want to slow down the cooling/warming process so everything has time to adjust.

makes sense i guess the blow torch method for taking apart lga775 chips is a no go?
07-24-2010, 05:19 PM
chew*

Quote:

Originally Posted by cdawall

makes sense i guess the blow torch method for taking apart lga775 chips is a no go?

well even if blowtorched, cutting the IHS seal is near impossible withoutnicking smd's. Add the fact that the core is fragile ( like 50% more fragile than deneb )chances are core will crush even if delid is sucessfull.
07-24-2010, 05:37 PM
Chumbucket843

i have never tried using LN2 to remove the die so idk if it's a good idea. you can use a blow torch to melt the solder bumps that connect the die to the pins or you can sand through it. i think the fumes when you do this are toxic so be careful. do it in a fumehood or outside b/c it stinks.also i have not deprocessed a cpu past p4 so i cant help get off the IHS on newer cpu's.

after this you need to do is epoxy the die to a hard flat surface then sand it with 2000 grit sandpaper. before you try doing it to cpu's practice sanding skills on DRAM's. it isnt going to be perfect because humans cant sand down with 1 micron of accuracy. a good way to test how well you are sanding is by how many levels you have exposed at once. if you are 3 or over be careful. look out for cracking. this will be your number one enemy because it will only get worse and bigger die are more prone.

i'm kind of confused with what happened to your thuban. it looks like you ripped the bottom silicon layer of the wafer off and the transistors and metal layers stayed soldered onto the packaging.
07-24-2010, 05:38 PM
cdawall

Quote:

Originally Posted by chew*

well even if blowtorched, cutting the IHS seal is near impossible withoutnicking smd's. Add the fact that the core is fragile ( like 50% more fragile than deneb )chances are core will crush even if delid is sucessfull.

i guess your right sanding down would work best. have you tested to see if there is large temp difference with a heavy lap vs none? and have you sucessfully sanded through a IHS yet?
07-24-2010, 05:59 PM
chew*

Quote:

Originally Posted by Chumbucket843

i have never tried using LN2 to remove the die so idk if it's a good idea. you can use a blow torch to melt the solder bumps that connect the die to the pins or you can sand through it. i think the fumes when you do this are toxic so be careful. do it in a fumehood or outside b/c it stinks.also i have not deprocessed a cpu past p4 so i cant help get off the IHS on newer cpu's.

after this you need to do is epoxy the die to a hard flat surface then sand it with 2000 grit sandpaper. before you try doing it to cpu's practice sanding skills on DRAM's. it isnt going to be perfect because humans cant sand down with 1 micron of accuracy. a good way to test how well you are sanding is by how many levels you have exposed at once. if you are 3 or over be careful. look out for cracking. this will be your number one enemy because it will only get worse and bigger die are more prone.

i'm kind of confused with what happened to your thuban. it looks like you ripped the bottom silicon layer of the wafer off and the transistors and metal layers stayed soldered onto the packaging.

I'm not confused. I hit it with the ban hammer 10 times for dying on me.

What you see is what was left after 10 blows with a 10lb hammer.
07-24-2010, 06:40 PM
cdawall

Quote:

Originally Posted by chew*

I'm not confused. I hit it with the ban hammer 10 times for dying on me.

What you see is what was left after 10 blows with a 10lb hammer.

well its obvious to me you have no idea what you are doing :rofl:
07-24-2010, 07:00 PM
chew*

Quote:

Originally Posted by cdawall

well its obvious to me you have no idea what you are doing :rofl:

Quite true if horse is dead get off and walk, no sense beating a dead horse err cpu....i highly doubt it will work any better now than it did prior to beating it. :p:
07-24-2010, 07:06 PM
richierich

love the money shot!!
07-24-2010, 07:33 PM
BeepBeep2

Quote:

Originally Posted by chew*

I'm not confused. I hit it with the ban hammer 10 times for dying on me.

What you see is what was left after 10 blows with a 10lb hammer.

LOL :clap:
07-24-2010, 07:39 PM
Sparky

So why is it more fragile than say an older 939 core? Like my Opteron 165 that I delidded? Which still runs fine BTW and I cranked down the heatsink on it good and tight, no issues, same with the guy who owns it now, never even chipped a corner :shrug:
07-24-2010, 07:46 PM
cdawall

Quote:

Originally Posted by chew*

Quite true if horse is dead get off and walk, no sense beating a dead horse err cpu....i highly doubt it will work any better now than it did prior to beating it. :p:

oh well some people don't appreciate what you do. thanks for killing the rest of your thuban/deneb's good luck with your clocking i wish i could join you like old times :wiggle:
07-24-2010, 08:08 PM
chew*

Quote:

Originally Posted by Sparky

So why is it more fragile than say an older 939 core? Like my Opteron 165 that I delidded? Which still runs fine BTW and I cranked down the heatsink on it good and tight, no issues, same with the guy who owns it now, never even chipped a corner :shrug:

it's even more fragile than deneb so knowing the only real change amd made was a low k dielectic.....I would say that it has something to do with that process.
07-24-2010, 08:13 PM
Serpentarius

i see so that's how the cpu looks like ... thx for sharing chew ... never had a chance to see it upclose
07-24-2010, 10:33 PM
Mad222

OMG...:eek:
What have you done, Chew*? A real CPU dissection!
This decent chip just brutally dismembered like this ?!
O..man. 1015 DPAW has become extinct:(
07-24-2010, 10:36 PM
crazydiamond

Quote:

Originally Posted by Mad222

O..man. 1015 DPAW has become extinct:(

just endangered :D

http://i70.photobucket.com/albums/i8...IMG_0765-1.jpg
07-24-2010, 11:02 PM
Mad222

Hey, Bro..that's a nice one
The number 313 in Chinese means "vivid & Healthy" :up:

Quote:

Originally Posted by crazydiamond

just endangered :D
07-25-2010, 12:57 AM
FatAlbert

I see imc on the photos chew :rofl:
07-25-2010, 03:10 AM
PatRaceTin

hmmmmmmmmmmm good!!!!!!!!!!!
07-25-2010, 09:08 AM
Tomasis

Quote:

Originally Posted by chew*

http://chew.ln2cooling.com/thuban/good%20shot%20.jpg

thats beautiful photo.. seems unreal..:eek::yepp::clap:
07-25-2010, 09:16 AM
chew*

Quote:

Originally Posted by Mad222

OMG...:eek:
What have you done, Chew*? A real CPU dissection!
This decent chip just brutally dismembered like this ?!
O..man. 1015 DPAW has become extinct:(

It sucked anyway, only good for wprime, and needed retarded amount of volts for it.

I lost my best 965BE to it as well.

My giga boards are whats extinct. I will wait for next rev boards,2.0 still has issues or at least mine did.
07-25-2010, 09:22 AM
Smartidiot89

Love your pics Chew* !

Quote:

Originally Posted by Mad222

OMG...:eek:
What have you done, Chew*? A real CPU dissection!
This decent chip just brutally dismembered like this ?!
O..man. 1015 DPAW has become extinct:(

I just recieved a 1015DPAW from AMD the other day, should I consider myself happy or is it just some lucky batch of yours?:D
07-25-2010, 09:40 AM
Mad222

WTH.....965be is gone too, damn!:shakes:
Problem with the power supply of ver. 2.0 ....it seems
While running wPrime in high volt, the board shuts down. It happens to both UD7 & UD5. The UD5 gave me 88 & C1 ...dead meat afterwards

Hope that problem resolved on next ver. 2.1

Quote:

Originally Posted by chew*

It sucked anyway, only good for wprime, and needed retarded amount of volts for it.

I lost my best 965BE to it as well.

My giga boards are whats extinct. I will wait for next rev boards,2.0 still has issues or at least mine did.

I envy you guys....able to get supports from AMD
We get nothing from them here in HK. We need to buy them on our own:mad:

Quote:

Originally Posted by Smartidiot89

Love your pics Chew* !

I just recieved a 1015DPAW from AMD the other day, should I consider myself happy or is it just some lucky batch of yours?:D
07-25-2010, 09:47 AM
chew*

Quote:

Originally Posted by Mad222

WTH.....965be is gone too, damn!:shakes:
Problem with the power supply of ver. 2.0 ....it seems
While running wPrime in high volt, the board shuts down. It happens to both UD7 & UD5. The UD5 gave me 88 & C1 ...dead meat afterwards

Hope that problem resolved on next ver. 2.1

I envy you guys....able to get supports from AMD
We get nothing from them here in HK. We need to buy them on our own:mad:

I thought the issue was OCP but it appears it is not.

Let me explain what happened to my board.

I benched my board fine in 32m and 1m.

Then we took the board and put a diff chip in. It was going fine. Then all of a sudden it just powered down. This was also during a session of 1m/32m which should not stress board much at all and we had cores disabled. Board is waterproof so it's not condensation.

After that one freak power down i broke it down. Board was never reliable since then. Upon inspection of mosfets there were no blow out so i suspect somewhere in the power deliver to fets is where it blew.

I can hear it click when i press power button by 8 pin connector but nobody is home.

Actually those guys are lucky, me i have to work for any support i get. I don't get paid to do 2 weeks prep for an event + the event......but i do get to keep a few cpus.

When i say 2 weeks i literally means straight working with little to no sleep. I literally slept walked/benched through black ops on maybe 30 hours of sleep in 2 weeks.
07-25-2010, 10:15 AM
Mad222

OCP was my 1st thought too
I had the same waterproof grease from TW you have... I believe
Exactly the same situation here...actually, 2 UD5 died....one was dead for no reason under cold & another one died with some parts burned but I was unable to find the exact location...somewhere around the mosfets

Nothing happened when running PI32m/1m on mine. But for wPrime 1024m, it just powers down like hell. Eventually, I freese the whole Northbridge HS & barely complete a wPrime run

As part-time overclockers, we did our best to participate in this contest & that's the spirit, Chew ! :clap:

Quote:

Originally Posted by chew*

I thought the issue was OCP but it appears it is not.

Let me explain what happened to my board.

I benched my board fine in 32m and 1m.

Then we took the board and put a diff chip in. It was going fine. Then all of a sudden it just powered down. This was also during a session of 1m/32m which should not stress board much at all and we had cores disabled. Board is waterproof so it's not condensation.

After that one freak power down i broke it down. Board was never reliable since then. Upon inspection of mosfets there were no blow out so i suspect somewhere in the power deliver to fets is where it blew.

I can hear it click when i press power button by 8 pin connector but nobody is home.

Actually those guys are lucky, me i have to work for any support i get. I don't get paid to do 2 weeks prep for an event + the event......but i do get to keep a few cpus.

When i say 2 weeks i literally means straight working with little to no sleep. I literally slept walked/benched through black ops on maybe 30 hours of sleep in 2 weeks.
07-25-2010, 10:17 AM
chew*

Quote:

Originally Posted by Mad222

OCP was my 1st thought too
I had the same waterproof grease from TW you have... I believe
Exactly the same situation here...actually, 2 UD5 died....one was dead for no reason under cold & another one died with some parts burned but I was unable to find the exact location...somewhere around the mosfets

Nothing happened when running PI32m/1m on mine. But for wPrime 1024m, it just powers down like hell. Eventually, I freese the whole Northbridge HS & barely complete a wPrime run

As part-time overclockers, we did our best to participate in this contest & that's the spirit, Chew ! :clap:

Is your board clicking near 8 pin when you try to power it on still?

I have a 890GX here that is dead on air, similar issues but reboot loops.

Tbh i only joined comp to change the game madd. Tired of seeing backed by manufacturer guys win hardware. I planned to give whatever i won to a guy further down list that can provide a reciept for there board and cpu. In my opinion the winners of hardware should be those that paid for all of there hardware.

You need to paint your boards madd. That grease is not 100% reliable. Let me take a pic of mine.
07-25-2010, 10:27 AM
Mad222

Yes....near the 8 pin !
890GX ?! Oh...tell me about it. I had mine dead too:mad:

Good for you...I think it's right to do :up:

I'm going to take a few days off next week & staying with my kids !:D

Quote:

Originally Posted by chew*

Is your board clicking near 8 pin when you try to power it on still?

I have a 890GX here that is dead on air, similar issues but reboot loops.

Tbh i only joined comp to change the game madd. Tired of seeing backed by manufacturer guys win hardware. I planned to give whatever i won to a guy further down list that can provide a reciept for there board and cpu. In my opinion the winners of hardware should be those that paid for all of there hardware.
07-25-2010, 10:37 AM
Smartidiot89

Quote:

Originally Posted by Mad222

I envy you guys....able to get supports from AMD
We get nothing from them here in HK. We need to buy them on our own:mad:

This is purely for review-purposes and the like, I am sure I would get nothing if it was for overclockning :rofl:
07-25-2010, 10:59 AM
Mad222

Hey, if it's for review purposes, nothing to concern with its' batch or oc performance then

Quote:

Originally Posted by Smartidiot89

This is purely for review-purposes and the like, I am sure I would get nothing if it was for overclockning :rofl:
07-25-2010, 11:02 AM
Smartidiot89

Quote:

Originally Posted by Mad222

Hey, if it's for review purposes, nothing to concern with its' batch or oc performance then

Needs to be tortured just a little after a thorough review lol
07-25-2010, 11:09 AM
chew*

here's the stuff i use and pics of PWM, as you can see no mosfets blew out and condensation can not effect anything on my boards.....they are just sealed from elements.

Only things I leave exposed are tops mosfets and nb/sb die....

http://chew.ln2cooling.com/890FXA-UD7/111.jpg

http://chew.ln2cooling.com/890FXA-UD7/222.jpg

A shot from a better angle to show how well everything is coated.

http://chew.ln2cooling.com/890FXA-UD7/333.jpg
07-25-2010, 11:26 AM
Daveburt714

Is it just me, or does that cap right above the 8pin connector (pic #2)
look like it's bulging?

Could just be the pic I guess... :shrug:
07-25-2010, 11:30 AM
Mad222

It's pretty well covered by the paint you used & shouldn't have any condensation problem
The burn must have been occurred somewhere around the 8pin.... very small click with light burning smell on my UD5
I supposed it was caused by overheating with full power on
Not sure whether it's the problem of Thuban or the initial design of 800 serie mobos
Well, I think MB manufacturers should be notified & aware of this incident
07-25-2010, 11:31 AM
chew*

Quote:

Originally Posted by Daveburt714

Is it just me, or does that cap right above the 8pin connector (pic #2)
look like it's bulging?

Could just be the pic I guess... :shrug:

nah its just my camera doing the fish eye.

Visual inspection even with a magnifying glass turns up nothing.

I got no smoke smell and if something did burn on my board the stuff i use is highly flammable.....board would go up in flames. Quickly.....

My guess is a trace burned in middle of PCB layers on mine.

One thing i did notice is swapping PSU's made a big diff in how much volts i could use and there was no rhyme or reason to it. I started with a single high amp corsair 850.....then swapped to thermaltake multirail 1200, then antec 1200 multirail, last was turbocool 1200 single rail.

I think the issue lies with whatever is delivering the power to the PWM. Some component in that connection is failing. Thats my best guess. I think PWM is more than capable of handling it, just the way the power is being fed to it has some issue.

There is one more alternative.....feedback. Sort of like a car audio system. you can send 1000 watts to a 500 wat speaker as long as the signal is "clean" but distorion in signal will blow speaker.....on the other hand amps don't blow speakers....distorion does.....if anything speakers blow amps with feedback......this could be the CPU not liking volts and sending feedback back through PWM.
07-25-2010, 12:34 PM
[XC] Oj101

OMW, that's amazing! I've seen pictures of cores before but always thought the colouring was an added effect, as they do for bacteria and so on for emphasis :shocked:
07-25-2010, 05:00 PM
Chumbucket843

Quote:

Originally Posted by Oj101

OMW, that's amazing! I've seen pictures of cores before but always thought the colouring was an added effect, as they do for bacteria and so on for emphasis :shocked:

they do that for "pro" die shots but it is still colorful with out it in most cases.
07-25-2010, 05:26 PM
BeepBeep2

hmm...
07-25-2010, 05:39 PM
Sorrow13

Thanks for the shots! Great pics. Sad the cpus died, but at least in death they can bring us joy.
07-25-2010, 06:02 PM
tbone8ty

i smell a wallpaper with this photo :p
07-26-2010, 04:10 AM
bartolomius

beautiful photos :D
07-26-2010, 06:42 AM
imamage

Nice pic you got there
So shiny !
07-26-2010, 01:15 PM
Tomasis

Quote:

Originally Posted by tbone8ty

i smell a wallpaper with this photo :p

better than burning smell :D

I hope Gigabyte reads this thread
07-26-2010, 01:57 PM
stunned_guy

Well done chew ;) Unbelievable pictures, the last one is great !
That remember me to my action with Phenom I X4 and Phenom II X4 taking the heat spreader of: http://www.hardwareluxx.de/community...te-706977.html
same fun and pics. enjoy.
07-26-2010, 04:09 PM
Chumbucket843

Quote:

Originally Posted by stunned_guy

Well done chew ;) Unbelievable pictures, the last one is great !
That remember me to my action with Phenom I X4 and Phenom II X4 taking the heat spreader of: http://www.hardwareluxx.de/community...te-706977.html
same fun and pics. enjoy.

just some tips:

the die is packaged facing down (it's called flip-chip) so you must sand in the other direction to see the good stuff. :fact:

haste makes waste :up:
07-26-2010, 11:35 PM
dafour

Nice,cool to see it naked.
07-27-2010, 12:32 AM
Don_Corleone

Phenomenal photos chew :clap: Specially the money shot
This click sound looks familiar for me. I heard this using BB XPower :confused:
In my point of view it related to OCP.
I just received my 890FXA-GD70 and gonna test it with a couple of wprime runs. Hope the same does not happen to my prolific 1090T.;)
07-28-2010, 02:21 AM
youngpro

hah brian thats sick, giving the people what they want as usual mate :)
07-28-2010, 07:25 AM
FlanK3r

no, i want no kill my chip :-)...nice photos and strange problem (only with boards 800 chipset? or 700 too?)
07-28-2010, 09:40 AM
[XC] Oj101

Hmmm, my dead chips normally get their pins removed and a hole drilled in order to use as a keyring, perhaps I should delid? :D
07-28-2010, 09:44 AM
BeepBeep2

Quote:

Originally Posted by Oj101

Hmmm, my dead chips normally get their pins removed and a hole drilled in order to use as a keyring, perhaps I should delid? :D

You should delid it and scrape the remaining rubber off :yepp:
07-28-2010, 12:02 PM
stangracin3

Quote:

Originally Posted by BeepBeep2

You should delid it and scrape the remaining rubber off :yepp:

if he is using it as a keychain, don't delid it...

i had an old celeron(working) that i did this to

i ground down trhe pins so it ended up dead, but after a few weeks, the core was DESTROYED.
before i made it a keychain the core was perfect.
if i can find it i will take a pic.
07-28-2010, 12:38 PM
=SOC= Admiral

Quote:

Originally Posted by 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.

http://chew.ln2cooling.com/890FXA-UD...lidded%201.jpg

http://chew.ln2cooling.com/890FXA-UD...lidded%202.jpg

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.

Now I wasn't doing anything for a competition but when my gigabyte UD5P went it took my first 720 with it and the one I got back was really go picky to oc.

All times are GMT -8. The time now is 03:42 AM.

XtremeSystems