agree with you owi :up:Quote:
Originally Posted by owikh84
Printable View
agree with you owi :up:Quote:
Originally Posted by owikh84
Ok. Delided my 3570k. I first tried some thermal fusion 400 while I waited for my liquid pro to come in. Got about 5c lower with the CM thermal fusion.
Now my Liquid Pro came in, I cleaned everything up put the LP on the die, and a super thin layer on the back side of the IHS.
WOW this is the solution for sure folks! I got >20c lower at load!
well i just got a new 3770K batch 3224B259...wich i can boot into windows at 5Ghz HT at 1,328 if i put my SLI with the Galaxy's GTX680 SOC needs more but look temps
IMC is great goes to 2933
this is 4.8HT 1.328 NO IHS REMOVED and mem 2800CL11 Room is 25C
http://i.imgur.com/3X1QY.jpg
My other chip was 98C !!! at 1,35v same ambient
i found the screen !!! OMG
http://i.imgur.com/Z9ujo.jpg
Cheers to all
Sergio
Nice chip Sergio! Do you think it can handle some linx at that voltage?
Good chip buddy,make sure not to damage it when removing ihs :)
Quote:
Originally Posted by McDown
thanks buds yes it can handleQuote:
Originally Posted by Alex-Ro
and i already removed his to 7 chips no damage, some ppl are having problems scratching PCB cauz the put on a flat table so they dont have the correct angle to put the blad, grabbing the chip with the hand is the best method
but this chip i show wont be removed i luv the temps and all keeper !!!
Cheers !!!
Sergio
Testing IMC 1415 Mhz - FULL BANK - @ 4.85 Ghz Batch 3226B987
http://www.firegoon.com/images/ay8j7b3.jpg
http://www.firegoon.com/images/f4jjrq7.jpg
http://valid.canardpc.com/show_oc.php?id=2591749
More testing --- 5.05 Ghz IMC 1413 FULL BANK
http://www.firegoon.com/images/7h1urc7.jpg
http://valid.canardpc.com/show_oc.php?id=2591765
I might have screwed up my i7 3770k while removing the IHS, the green pcb outside of the IHS, the blade chopped off some green stuff from surface of the cpu, it shows up some copper.
Channel A1 and A2 memory slots wont work but otherwise cpu works fine (error 55, memory not installed when i put memory to A1 and A2 slots), need figure out if i can still fix this cpu.
And there are some slightly bent pins below same spot in socket so it could be that it didnt damage at all but pins just dont make contact.
That's the risk we all run...
Look at these pictures are of an 3770k killed in battle ( mine not, from spanish forum )
http://www.firegoon.com/images/r8sraf8.jpg
http://www.firegoon.com/images/64eiabi.jpg
Original thread: http://foro.noticias3d.com/vbulletin...391334&page=68
^^Sorry, but that's a stupid way to waste a good working cpu....Too bad not even go cold yet:(
Probably spanish guy used a machete...:D
Lucky i only broke Memory Channel A from CPU, i tested the cpu otherwise and apart from Memory Channel A its working without problems.
So memory bandwidth is bit lacky at 15500-16600Mb/s when it could be double that, but at least this IMC on the cpu can do something like 2400Mhz with my G.Skill sticks that are rated for 11-11-11-31-T2, but im running them at 10-11-10-28-T1, maybe could go to CL9 without problem.
http://imageshack.us/a/img809/844/hy...ngechan.th.pnghttp://imageshack.us/a/img820/6681/a...nelddr3.th.pnghttp://imageshack.us/a/img820/1898/c...0k49ghz.th.pnghttp://imageshack.us/a/img526/1315/s...nglecha.th.png
Bit slow at HyperPi tough :/
lol!Quote:
Originally Posted by brama
I used "utility knife" or "stanley knife" at start but soon as i noticed it started to go wrong i switched to thin razorblade, should have used razorblade from start so there would have not been damage to cpu.
:rofl:Quote:
Originally Posted by Brama
This is one wierd cpu, running 1.40v at 4.8Ghz is faster in Cinebench and LinX than 5Ghz at 1.5v and that 5Ghz is at least 2hours LinX stable.
I have a QX6700 which is missing a small part broken off the green substrate but it's at the edge. Some copper could be seen as well. Has been running for 3 years already, the damage was already there when I got that CPU.Quote:
Originally Posted by rintamarotta
Yeah, i think it will run fine till i deside to send it to garbage bin. Going to buy two other i7 3770k's anyway in next week, so i think ill just do suecide run with this cpu, 1.8-2.0v and 5.5Ghz in phase change when other i7's come in from my mailbox.
Or then im going to fire up my soldering iron and actually solder those points where copper is visible and then test what happens, need just get some cheap Z77/Z68 motherboard to sacrifice for this cpu.
when highers speeds produce lower results than lower speeds means that something aint stable...Quote:
Originally Posted by rintamarotta
Probly because of no retention bracket used on mobo's socket cpu wasn't make full contact with socket pinsQuote:
Originally Posted by rintamarotta
All pins make contact, its clearly visible from pads in bottom of CPU. Damage is in green pcb of cpu, there is two places where copper is visible and i think one of the visible copper ones have disconnected "wire" going to socket pin inside the cpu pcb so i might need just fix it by using copper conductive pen to reconnect the "wire".Quote:
Originally Posted by Dumo
And i use retention bracket to hold down cpu, i just put in proper thermal paste betwean cpu and ihs.
Gave my second 3570K a try since i failed and only single channel working on the first one. This one works 100% and with custom cooling i?ll get around 60 degrees with 100% load and 4.5ghz with 1.3v. I?m happy :) Before with stock i got 81 at the highest core. Using LP on the die and mx4 between IHS and Cpublock
less volts with new bios 1408 NO IHS removed and 25C ambient room summer
running
http://i.imgur.com/y14wO.jpg
done
http://i.imgur.com/fAiEh.jpg
3226B987
4.3 Ghz 1.088 Vcore 1401 Mhz FULL BANK
http://www.firegoon.com/images/1797crr.jpg
4.8 Ghz 1.27 Vcore 1401 Mhz FULL BANK
http://www.firegoon.com/images/rfr3quj.jpg
My 24/7 for summer time
http://www.firegoon.com/images/7rjwcww.jpg
Getting ready to take the razer to my first new 3770k.
Final testing before delidding..
Batch 3226C840 I7-3770k
3770k@4.5GHZ vCore 1.15v
4.5GHZ
Watercooleed with UT-60 420mm RAD and DT Sniper water block.
TIM IC Diamond
Not Delidded
Ambient Temps 21C
http://valid.canardpc.com/2595435
HT on
http://minidriven.com/BlacknBlueforI...4p5HTonIBT.jpg
Package Min 26c Max 66C
HT off
http://minidriven.com/BlacknBlueforI...aseIBTtest.jpg
Package Min 23C Max 66C
Time to see if a little delidding magic will happen. Wish me luck! :cool:
The deed is done. My first 3770k delid. This one was much tougher than my 3570k. Lid was tight. Worked 2 different razers to get the lid off.
http://minidriven.com/BlacknBlueforI...3770ksmall.jpg
http://minidriven.com/BlacknBlueforI...3770ksmall.jpg
Bare Die Mount no IHS.
http://minidriven.com/BlacknBlueforI...mountsmall.jpg
Good job,share us the results :)
Quote:
Originally Posted by SonDa5
After delidding.
Ambient Temps 22C
Hyperthreading On
http://minidriven.com/BlacknBlueforI...s3770kHTon.jpg
Hyperthreading Off
http://minidriven.com/BlacknBlueforI...3770kHToff.jpg
Here is 4GHZ Intel Burn Test Run on my water cooled delidded 3770k with Hyper Threading enabled.
Ambient temps 24C
1.09v on vcore, 1.5v PLL
TIM Coolaboratory Liquid Pro
Direct Die to water block mount. No IHS with DT Sniper water block.
http://valid.canardpc.com/2597882
http://minidriven.com/BlacknBlueforI...age3770IBT.jpg
Nice work Sonda5 now Try 0.98 Vcore and put 1.40vcore on CPU PLL
3226B987
http://www.firegoon.com/images/vm6cbxh.jpg
im really annoyed with my 3770k i really wanted to hit 4.5ghz as it seems the norm but i have done ihs mod hoping it will help with my temps but its made no difference with my temps at all im using Arctic Cooling MX-4 on the die and the h60 cpu cooler the other thing is i need 1.3v in bios to keep it stable anything less and it freezes or bsods plus it hits 103c is there something im doing wrong or have i got a really s++t cpu
thanks
rubber5 : You must use ( liquid metal ) coollaboratory liquid pro / ultra or Phoebya LM...is the only way that you can downgrade 30 to 40 degrees
forget it arctic mx or silver or gelid..use Liquid metal!
Quote:
Originally Posted by hersounds
Will try that later on. Will also set my ram to your settings from your screen shot.
Sounds like you have a bad mount. Probably poor contact some where between the IHS and DIE or IHS and heat sink surface.Quote:
Originally Posted by rubber5
Vcore at 1.2v. PLL at 1.55 and you should be good to go. You have a cooling problem.
1.4, 1.45 no post on PLL. 1.5 PLL good.Quote:
Originally Posted by hersounds
Batch 3226C840
http://minidriven.com/BlacknBlueforI...hz3770kIBT.jpg
http://valid.canardpc.com/2598378
U have atleast a good cpu with low voltage, my 3570K is barly stable with 1.3v at 4500mhz
That's perfect Sonda5...you have same 3226 and is a nice batch
I just got in the same batch chip yesterday, on water so far it can do 4800 at 1.325v and 5G at 1.45v though imc is just average but still a decent chip.
Now i delidded my third 3570K, using Liquid Pro on the die then on the IHS i using MX4. Is MX4 the best out there? And if i would like to run without IHS , shouldn?t it be good with a thermal gap filler then. Feels so easy that i?ll crush the die.
i reapplied but with artic silver 5 i used the spread method under the ihs and the line under the cpu cooler i have gaind 10c on my hottest core and now im at 4.5ghz with the 3770k @94c but it was at 104c and it did start to throttleQuote:
Originally Posted by hersounds
i have also seen that prime 95 v27.3 blend and ibt gives me the same temps at the end of the day thats strange i thought ibt was ment to be 10c higher
anyway im happy with my oc now
its just a shame i have 1.32v in bios llc set to extreme to keep it stable even dropping it to 1.31 gives a bsod :(
when i was using 1.3 it was fine but after tying again after reapplying the paste it did bsod on me
im not sure if it worth using liquid metal to help im a bit nervous using it becuse of it been able to damage some cpu coolers becuse of the cemical reaction
thanks for your help tho
Try lapping the bottom of the IHS a bit, the part that sits on the pcb. That will get you a tighter mount. The liquid ultra just works better with bad mounts which is what you and I had. Its the reason that some people see such a large improvement with using it but there are other workarounds. Lapping the base of the IHS and doing a really good job cleaning up the epoxy is what did the trick for me.
Yes the big problem is the black epoxy you must delete from pcb
http://www.firegoon.com/images/tqgjk7w.jpg
...and of course a little lapping help so much
http://www.firegoon.com/images/ba7mhth.jpg
I just tried delidding my 3770k by the lid is on incredibly tight. I managed to get the razor underneath, but then this happend. :(
How screwed do you think I am? :confused:
http://i.imgur.com/lewzm.jpg
http://i.imgur.com/Zfj96.png
before (pre-ihs removal)
http://i.imgur.com/GUrkT.png
during
http://i.imgur.com/8bZrn.jpg
liquid pro applied (4 surfaces - cpu die, both sides of ihs, waterblock)
http://i.imgur.com/PxORH.jpg
after
http://i.imgur.com/XTJhg.png
http://i.imgur.com/xfQjX.png
= 35c load temperature drop (104c to 69c)
That is an awesome pic, the horror inside each chip is real. Liquid pro works extremely well yes.
Crazy.Quote:
35c load temperature drop
Did you remove the IHS, put the liquid pro, then put the IHS back on?
when i done mine i could see a little metal showings it was scrapings of the ihs just wipe it with tim clean it looks ok to me just put it in your rig and try it thats the only way to tellQuote:
Originally Posted by barretp
You could try covering any exposed metal/traces with something maybe nail varnish or something if you think there's a danger of some other metal like the IHS or LP contacting the exposed area.
:ROTF: :rofl:Quote:
Originally Posted by Seeker-
Any benefits from lapping the die of the CPU directly, andressergio? I noticed a few others doing it recently.
Also has anyone tried placing the CPU back into the socket without the IHS? Do the brackets surrounding the socket & latch sit above, flush or under the die?
I'm considering running the chip without the IHS and having the die contact the waterblock directly. From the looks of things, the sides of the CPU will need some support in the form of a plastic washer or something due to the metal pressing down on the left/right of the chip if I were to leave the brackets in (I'm not sure if you can do away with them due to the pressure they put on the pins and such!)
Seeker-: imho if you use LM as TIM1, then there are almost no gains in lapping die itself or underside of IHS. Also if using IX or LM as TIM2 between IHS & cooling system.
I disagree. I used LMP (and LMU) on my de-lidded 3770k and after lapping the water block and IHS temps dropped a further 6-10 degrees C under P95 blend and IBT Max.Quote:
Originally Posted by Church
I am talking about removing the IHS completely and lapping the die of the cpu. I know some people will say there aren't any gains to be had, but I can't see it making anything worse as long as the entire die contacts the waterblock.
I'm just not sure if I can do it without needing to remove the entire latch and bracket system surrounding the socket.
if you lapping the die...sillicon...the chip die in the first move...forget it!
Good news everyone, I did not kill me chip. I got my newly built system running last night. I'll post some pictures after I get home from work today.
Question about Liquid Metal Ultra: what's the best and/or easiest way to remove the LMU TIM for a new application?
I used it between the die and IHS....I had to pull the cpu from the board to test something else....now it's time to re-install.
Edit: Wiping with cheap toilet paper and then 91% isopropyl worked fine.
Just Unlidded and what a difference in temps and stability , Im so impressed i decided to start my clocking from beginning again and see how low i can drop Vcore and temps with a set of 2133 memory and once I hit bottom Ill work back up to a good power to heat ratio . So far everything has been successful with a 25c drop in celcuis in prime of 3 hours and IBT passed. I was able to get IHS off with out a scratch and my top was a hella tight from my water block . but was well worth the drop in temps and vcore I'm achieving , its killing my best sandy s by a lot . Right now I'm priming as Im doing this 4.7 @ 1.330 Vcore
Top Block: Phobya HeGrease Extreme Thermal Paste
Under Top: Coollaboratory Liquid Ultra Paste
http://i84.photobucket.com/albums/k7...0kUnlidded.jpg
http://i84.photobucket.com/albums/k7...kUnlidded1.jpg
http://i84.photobucket.com/albums/k7...kUnlidded2.jpg
http://i84.photobucket.com/albums/k7...me1330pass.png
http://i84.photobucket.com/albums/k7...ed1330pass.png
The stuff works incredibly well when done right. I did a 3227C that runs retarded cool in my Torture rack system on water.
Thumb up for cleanliness & awesome temps!Quote:
Originally Posted by the_real_7
Thanks owikh84 , was a bit scary taking the jump :eek: but well worth it ! ! !Quote:
Originally Posted by owikh84
whit org IHS on
4.5GHZ
http://img7.imageshack.us/img7/6156/testtempcpu.jpg
5GHZ
http://img441.imageshack.us/img441/7983/temp25.jpg
http://img805.imageshack.us/img805/5694/img0896q.jpg
how many have run 5.2 or 5.4GHZ after they change's CPU,whit new pasta ???????????????
I'm sorry for my mess sys :D
The last two I have tested went 5100 and 5250, my current 3227Cxxx bare can do 5300 though it takes some juice. All binning is done on my torture rack water system.
is it whit 3D mark ? you can run 5.2/5.3GHZ
Yes I usually use FM11 or Vantage and 3DM06.
wow :up:
Here is a quick run of FM11 at 5150mhz with a 680 on water done last week. http://www.3dmark.com/3dm11/5148895:cool:
Your CPU score is very low,I get 1000 more than you by 5GHZ
i use a Maximus IV Extreme P67
http://hwbot.org/submission/2332765_
Hi hersounds, I am new to PC building but I followed everyone here on this thread closely now, I am getting a chip after vacation (Christmas) and I'm gonna do this. Did you get better temperature using Coolaboratory Liquid Pro vs. Phobya LM?
Thanks man and everyone else here for such great results.
Also my cooler is Corsair H100i with copper plating, should I use Liquid Pro between IHS and copper heatsink or more normal TIM material?
Thanks!!!
Welcome to xtreme Systems...Yes coollaboratory works a little better that Phobya LM,about corsair H100I is not problem you can use between IHS and copper heatsink.
be careful with metal liquid, take really care, and don't start your system after be safe that the work its fine as well.
NO,do not use liquid pro in contact with copper-base or aluminium.Also do not use on ihs !!!! You can use liquid metals ONLY on naked cpu-s and cpu coolers with nickel plated base such as prolimatech megahalems for example.
http://www.frostytech.com/articleima...alems_det2.jpg
Gallium disolve the aluminum but not copper alex...use just a little and relax nothing to worry...just " lapping" a little or use isopropyl alcohol to clean and that is all.
using just a little does not stick or weld
Thanks for the welcome. :)
I see there are some different response for Liquid Metal stuff + Copper heatsink + IHS top.
I know Liquid Metal uses Gallium, and Gallium kills Aluminium. I will research Gallium/liquid metal vs. Copper now! And find out more information...
BTW guys, I found this. It was interesting! I'm sure you all heard of this "Indigo Xtreme" TIM before, it's a bit different. Anyway this guy said he tested with a bunch of different TIMs, including IX. Delid CPU, apply Liquid Pro to the die, then test with different TIMs.
http://www.overclock.net/t/1260540/i...mx4-pk1-lqp-ix
I hope this is helpful and maybe more interesting information for you guys!
Cheers
NO it attacks copper too,trust me,i made the research before using it ;)Quote:
Originally Posted by hersounds
I will look attention to that ... so my block is in serious danger from a few months ago ? I don't see anything wrong on cooper base, but i repeat i will look attention
Used it on a copper waterblock for 4 years.Quote:
Originally Posted by Alex-Ro
When I removed it, there was nothing but a slight residue left over on each part.
Cleaned it off, and it was good as new.
Ihs Mod #2 3770K
3229B068 - Costa Rica
First ... knowing the edges ...
http://www.firegoon.com/images/akv539d.jpg
Next...Loosening IHS ... relaxed.
http://www.firegoon.com/images/7jfwupc.jpg
Welcome to the die!
http://www.firegoon.com/images/tr3f7k5.jpg
TIM crappy foreground, much crap to clean.
http://www.firegoon.com/images/ejdu6ey.jpg
Alcohol, swabs and liquid metal ready for action.
http://www.firegoon.com/images/debhidw.jpg
Chip absolutely clean and undamaged, next stage a mini lapping the IHS inside and out
http://www.firegoon.com/images/xrvcx2j.jpg
Mini lapping inside, black rubber removed to get more pressure with the same micro die.
http://www.firegoon.com/images/f1e2btt.jpg
Mini lapping outside, outside, to get the best dissipation of the liquid metal
http://www.firegoon.com/images/e7gbkhx.jpg
Coollaboratory inside the inner ihs + Pro Coollaboratory on the silicon surface.
http://www.firegoon.com/images/5pqw1hi.jpg
Cool laboratory Pro sobre el IHS del chip , ya cerrado listo para bootear y poner en uso.
http://www.firegoon.com/images/6n3m54t.jpg
Next stage...tested!
Waiting for results!!!! :D
First testing 4.8 - 1.24 vcore successful
http://valid.canardpc.com/2631625
http://www.firegoon.com/images/p2tg5cj.jpg
4.0 Ghz @ 0.96 Vcore 26C ambient
http://www.firegoon.com/images/wv1335n.jpg
Went to Micro center this morning and they had more 3032b chips which I grabbed another as my first one can do 5G at 1.34 and 5500 max. They have at least 6 more 3032b368 chips in stock.
Che hersounds, very nice buddy ;)
My 3770k died a honorable death, but I'm getting a replacement soonish. I'll delid and post some results here...
I already got Liquid Pro and that same sanicol bottle lol
Feliz a?o!
Is it better to put Liquid Pro on ALL INSIDE of the IHS? And also on silicone of die? Like you did on this new test?
Or just put some on the silicone of die only?
Seems A LOT of Liquid Pro if you put all on the inside of IHS.
Thanks!
Just a word of warning I'd rather be told I'm wrong than see someone mess up but don't use that much LP.
You only need a tiny bit & spread it out thin, using too much may run the risk of some of it running/dripping off the die & could potentially short a trace or drip out the gap in the IHS & damage/short something else on the board/video card.
Small half rice grain drop on the ihs and spread it, and similar drop on the underside of the ihs and let it spread with the mounting pressure...for the top of the ihs LP is a bit evil cause it's hard to remove and erases all processor's markings, but it's optimal, specially if you lap it.
thanks buddy, sorry for your chip... y excelente feliz a?o nuevo! ( happy new year )Quote:
Originally Posted by Ivanlabrie
Yes is better put in 4 sides sillicon die - ihs inside - outside and block cooper.Quote:
Originally Posted by Deceit
perhaps the pics is like a lot of liquid metal but is only a little dispersed,the leftovers, was absorbed by the syringe.
yes this is right...but don't forget that the leftovers, was absorbed by the syringe...and the pressure of socket retention.Quote:
Originally Posted by C-N
YES.Quote:
Originally Posted by Ivanlabrie
one of reason i like WC, idle about 26C, Load @41C, 4.6ghz
Carelessness killed a 3770K today.
I was expecting the black gunk holding on the IHS to be tough to cut through; on my CPU it was not. The blade went right through and nicked the corner of the die pretty good. Stupidly I never tested the CPU before removing the IHS, but i'm pretty positive this 3770K is dead. From NiB to killed in under 5 minutes.
Be careful when starting to cut into the corner of the black adhesive, it might not be as tough as expected:
http://i399.photobucket.com/albums/p...ps7ccd36c6.jpg
http://i399.photobucket.com/albums/p...ps1cda37af.jpg
Yes I have done that, makes a rather expensive key chain. Take your time or pay the price.
I have one chip like this, it's still working fineQuote:
Originally Posted by trn
http://i458.photobucket.com/albums/q...4/06473f8c.jpg
I'm now the shameful owner of 2 killed 3770k's; I give up.
trn: two in row? what happened in 2nd round?
2nd CPU is on the right, 1st is on the left.Quote:
Originally Posted by Church
I'm not really sure. I was rough with the brillo-pad when removing the left over black IHS-adhesive; I might have damaged something then. There is also one tiny area on the upper left where the green coating is all worn off and copper is showing, I don't see damage to the copper though. I tried re-mounting this about half a dozen times and still no luck.
http://i399.photobucket.com/albums/p...ps2fda105d.jpg
http://i399.photobucket.com/albums/p...ps0182b73e.jpg
http://i399.photobucket.com/albums/p...ps41cdc7a9.jpg
mate, saw those at ocn too...why would you LAP your die and pcb like that???
Next time use your fingernails for removing the glue.
A Brillo pad? Why on earth did you use that? Fingernails or a creditcard are enough to remove the glue.
I use to use Brillo to remove all matter of crud off Copper/Stainless/Iron plate when I was young. Now using a Brillo for an Intel cpu, isn't that kind of going over the top?Quote:
Originally Posted by Zeus
I would use a dab of Mineral Turps to loosen the black-glue/sealant and then clean it all up with Alcohol. Is there any other way to clean the pcb other than scratching it with any object, ie, like your finger?
Doing it just with your fingernail or a creditcard is easy enough. It's not like you have to rub for hours. Got it done in less than 5 minutes. I'd rather not use any dissolvents on the pcb if i can do without.
Just thought I'd share my findings...... done the roof chop to a new 3770K I got at the weekend as it was running so hot @ 4.5GHz I couldn’t watch..
Before the chop the chip could run IBT @ DDR 2400 1.16vcore 85oC but failed @ 2666, more vcore only made things worse I suspect the chip didn’t like the temp more than the volts. After the chop temps dropped just over 20oC & now the chip is IBT stable & has been running prime torture for several hours.
I know the chip is a bit of a chunk blower but loosing 20+oC has made a huge difference to its stability & ability to clock memory.
frequencies and voltages are really affected by heat mainly in Ivybridge
post some 5GHz HT luv mate :DQuote:
Originally Posted by hersounds
I can not with these ambient temperatures and the antec 620 is not enough for cool 5Ghz ...
...but someday more cold, I can try it.