Attention Dead Micron D9 Owners

[Nate P.]

I just took out my G.Skill HZ sticks, still a no go.

[Ket]

I'm not sure how long my D9s have been in the freezer now, maybe 9-10hrs. I took the heatspreaders off this kit of D9s a long time ago, so maybe I'll have more luck with this "life" technique. If people are willing to do so where their not having any luck it would probably be best to remove the heatspreaders.

[Vivi]

well, my GTS 512 is half broken due to the memory, maybe this will kickstart it back to life? if it does THEN HOLYZHIT il be HAPPy.
the one in my sig lol.
it works in 2d mode, but as soon as drivers kick in its a no go

[nullface]

Hm, my ballistix are going in the freezer, they wont boot at anything below 2.3V anymore.

[OldChap]

One set of D9GMH'd Team 6400 4,4,4,10. just gone in the freezer. I'm game for a try even though it seems so unlikely, as has been said "nothing to loose"

[fatguy1992]

I'm going 2 blow up some of my old DDR ram and see if the freezer will fix it

[Reinvented]

Just out of curiosity, how much vaseline did you use? I have 3 sets of G.Skill HZ's. One which died, and the other two still work perfectly fine (4GB's), but I'd like to get that last set to work. They were the nice black PCB ones, that clocks super well.

I would definitely like to try this, and put it in the freezer. But also...how cold is your freezer?

[zsamz_]

Just out of curiosity, how much vaseline did you use? I have 3 sets of G.Skill HZ's. One which died, and the other two still work perfectly fine (4GB's), but I'd like to get that last set to work. They were the nice black PCB ones, that clocks super well.

I would definitely like to try this, and put it in the freezer. But also...how cold is your freezer?

put it in a ziplock bag no vaseline
use a blowdrier on it to make sure its dry n let it sit a while

[HDCHOPPER]

I have a dead set of D9s, I'll give this a bash too

if it doesnt work you may want to put them back on the mobo and warm them up first

the stick eye recovered were all fresh off the mobo when chucked into the freezer

[[XC] Oj101]

Just out of curiosity, how much vaseline did you use? I have 3 sets of G.Skill HZ's. One which died, and the other two still work perfectly fine (4GB's), but I'd like to get that last set to work. They were the nice black PCB ones, that clocks super well.

I would definitely like to try this, and put it in the freezer. But also...how cold is your freezer?

I used just enough to seal the gap between the PCB and the heatspreaders. If you leave it to dry thoroughly afterwards there's no need for insulation.

It's just a normal fridge freezer, I'm guessing about -10'c?

[Ket]

home freezers are pretty potent. I'd say an average freezer is about -18c, with good ones being -25c

ED - Just taken my D9s out the freezer, will stick them in in a few hours or so

[Xenogias]

I heard of something called cryo-annealing. where basically things are put into a liquid nitrogen bath to make them harder or something. Maybe this is a similar concept on a lesser scale. Also, I remember something about someone dunking a cpu in LN2 for a while to se if it would affect its OC-ability but it didn't. I dont think its been tried on dead cpus though. At the next LN2 party, someone should try dropping in some dead components =]

[CryptiK]

The average freezer is -20*C +/- 5*C. I have heard someone revived a dead cpu using this method, but I guess it really depends, like the ram, on what exactly is damaged and probably a lot of other factors, hence the less than 100% success rate.

[[XC] Oj101]

All the same, 75% success is huge

[Gautam]

Cut it out with the inter-forum crap. We don't need that here, thread pruned and infractions to follow.

[I34z1k]

Eh dunno what I missed but anyway

Gotta say, this has been happening for a while, amongst quite a few SA guys. I got 18mhz extra @ 2.1v with my crucials Exact same settings, back on my DFI 965s, yes, it can also be done for a slight gain in clocks. I've had some cases where I gained nothing, others where I gained quite a bit. In fact i froze a set of OCZ sticks in dice over the weekend and they're booting @ 1.8v again, after not being able to cause I ran them for a few hours @ 2.9v or so.

[N3RO]

One of my sticks gives lot of errors in Windows.

Just put the stick into the fridge... lets wait

[BiFfMaN]

The average freezer is -20*C +/- 5*C. I have heard someone revived a dead cpu using this method, but I guess it really depends, like the ram, on what exactly is damaged and probably a lot of other factors, hence the less than 100% success rate.

I agree, But still..just tossing it in the freezer does not cost you anything but time.

I would try it but i dont have any Dead Micron Owners to toss in the freezer...and besides its full of Hookers anyways.

[VIctorj]

Gonna try that tonight guys...report back probably a 5am local.

[[XC] Oj101]

Just to let you know, it's still working.

[saaya]

there are different packaging techniques for every silicon chip out there, memory, cpu, chipset, gpu etc...

every chip is usually so tiny that you can NOT directly put it on the pcb of the mainboard, memory module, vga etc... so you need an extra layer in between. so you have the silicon die that is connected to a tiny pcb, which has relatively large solder balls on the bottom, which can then be connected to the large pcb. to protect the silicon chip many packaging techniques submerge it in some sort of black ceramic, which is what we all know memory chips as. small black squares. well the actual chip sits inside there. cpus used to be submerged in ceramics as well, so were chipsets and gpus... but the heat properties of the ceramic arent that good, so the hotter a chip runs, the more it makes sense to have direct contact to the silicon die.

here are the main packaging techniques im aware of:

• there are actually silicon chips that can directly be soldered to the large pcb (uncommon as those solder balls are relatively large and a chip usually needs a lot more pins per surface than it can get with those relatively large solder balls)
• there are silicon chips that have tiny bga balls on the bottom, which touch a tiny pcb, which has normal bga balls on the bottom which then sits on the main pcb.
• there are silicon chips that are wire "bonded" to the tiny pcb which is then soldered onto the memory pcb.

the latter is the original technique and the most common one for memory chips.
then there are variations of that technique:

• you can solder the wires on the silicon chip and solder them onto the tiny memory chip pcb
• you can solder the wires onto the silicon chip and then basically smash the cold wire onto the tiny memory chip pcb and pretty much bend it on the contact pad
• you can smash the wires onto the silicon chip and smash them onto the tiny memory chip pcb.

the wires are tiny, so you dont need that much force to smash the wires onto the contact points on the silicon and on the pcb contact points. i dont remember which technique micron uses, but it involves smashing the wire onto a contact point on either the silicon on pcb side.
it sounds like both heating or cooling the chips can somehow fix wire bonding issues... how exactly im not sure... the metal bonds should stretch or shrink most in the whole package, the silicon and cermamic shouldnt stretch ot shrink as an effect of the temperature.
so it almost sounds like the wires came lose or lost good contact at least, and changing the temperature will rub the metal wires against the contact pads or at least increase the pressure of the wires against the contact pads, and thus improce the contact again...

very interesting!

[Jonny5isalivetm]

Hi Folks

This is pretty amazing to read!!



I have a set of Gskill HZ 2 x 1GB and after I go to bed the next morning the Computer wont boot so I have to reset the Bios etc..

Once they have warmed up awhile they do 600 5-5-5-15! tho currently at 2.55v lol

If I were to freeze these sticks you think it would help any ?

J5

[jas420221]

Tried on some of my old ballistx tracers, no go.

[[XC] Oj101]

Ah pity. Perhaps try longer?

[Revv23]

I'm giving this a shot on my HZ's.

It saved my mushkin BH-5 back in the day; though I highly doubt this will work on these sticks - as they still work just not without 2.55v