Jump to content
HWBOT Community Forums

Recommended Posts

Posted (edited)

Just wanted to check who's having problems with the tiny PCB of Skylake CPUs bending in the socket after delid?

 

Some folks are talking about it on Facebook ... check the picture below. Anyone else with similar experience?

 

attachment.php?attachmentid=3316&stc=1&d=1442986942

 

attachment.php?attachmentid=3318&stc=1&d=1442987284

Edited by Massman
  • Replies 51
  • Created
  • Last Reply

Top Posters In This Topic

Posted (edited)
The PCB is only 0.780 mm thick so it will bend without IHS if you put it back in the socket. just too many pins and the PCB is not thick enough for that.

 

This one has never been used without IHS Roman. ;)

Edited by pepinorang
Posted
This one has never been used without IHS Roman. ;)

 

Sure, but normally the IHS will keep it straight and/or the socket frame will fix it. So not really a problem for normal people :D

Posted (edited)
Sure, but normally the IHS will keep it straight and/or the socket frame will fix it. So not really a problem for normal people :D

Oh, yes agreed! Though Haswell CPUs won't bend if you accidentaly drop them from 50-60cm height on the floor while Skylake will bend real bad (which I also agree should not happen in the first place)...

Edited by pepinorang
Posted (edited)
dead link.

 

this is really convincing me not to delid a cpu.

 

This:

attachment.php?attachmentid=3322&stc=1&d=1443064122

 

It's from Fugger's FB: "6700K Retail popped" he also commented "Trying to push it too hard, 1.9v @ 6.2Ghz XTU goes pop -155c" :(

Edited by pepinorang
Posted

My delidded 4690k had warped so bad it wouldn't even post. I just used a makeshift jig and turned cpu upside down and very slowly heated it up with blow dryer and heat gun. Once straight I re-lidded with Kryonaut and lots of black silicone to keep IHS from separating from pcb and regain it's rigidity.

 

Wonder if we can fashion some sort of brace...
  • Crew
Posted
Any idea how to prevent it?

 

dont use a box cutter, it has a bevel which forces the PCB down when you slice it in.

 

WRONG:

non-segmented-small-cutter-blade-rb-07-429.jpg

 

 

use a minora blade

RIGHT

ae5890.JPG

Posted (edited)

Yesterday evening I wanted to push Cinebench some more to improve my score.

However, CPU refused to take more than 6.1G while it could do 6.3G before...

 

After unmounting (mounting was good), I noticed my chip was slightly more bent than before I start to bench.

I didn't try it after session and left it like this until now.

I just did and it's f***ing DEAD! #Bendlake :(

 

I think it's thermal related, I will probably avoid torching pot on this gen.

 

EDIT: And maybe try to glue back IHS too, I was just paranoid that paste would be contaminated by moisture over time...

 

attachment.php?attachmentid=3325&stc=1&d=1443171077

Edited by pepinorang
Posted (edited)
you heated up afer the session? to positive temps ?

 

I torched from -170 all the way to -20, then run XTU until ~60°C in XTU

I think I will now use heatgun for CB and unmount pot cold, let ambient temp heat up CPU.

Edited by pepinorang
  • Crew
Posted

my 4770k lost little bit of mhz every session after i glue it back.

 

i would stick to fresh paste each session, but be gentle guys!!! cpu's super fragile physically. maybe get your wife to delid and mount

  • Crew
Posted (edited)

maybe the heatup is too fast, not realy fond of this architecture. The fragility scares me. Most CPU's die with either the delid or after a perfect fine session... Especially the latter is like wUUUUUUUUUUUUUUUUUUUT happened here?

Edited by Leeghoofd
Posted (edited)
my 4770k lost little bit of mhz every session after i glue it back.

 

i would stick to fresh paste each session, but be gentle guys!!! cpu's super fragile physically. maybe get your wife to delid and mount

 

Well, on 23 chips delided there was very little casualties (2 ?), but it was done by professionals aka Leeghoofd, Zzolio, Wizerty and Strat :D

So far I only did two myself but they also survived without bending.

Edited by pepinorang
Posted

A big guy just told me that he never torch more than -40, then use XTU to +20 and never had issues with PCB bending so it's probably just my fault. :(

However, that kind of s*** never happened to me on 5960X...

Guest TheMadDutchDude
Posted

That's because the 5960X has a much thicker PCB. ;)

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...