[gwar2288]

http://outofspecs.gr/forum/showthread.php?t=3094

Post #12
The relevant Pictures of a Evga P55 FTW (You know, the one with the Holier than Thou LOTES socket) and a i7 860.

[EarthDog]

Interesting thread title.....why the aggresion?

Good find, though I cant get it to translate at all. First Lotes socket I have seen cooked! Though it appears its up at 1.57 - 1.64v as well...extreme cooling.

EDIT: But later in the thread AFTER it was 'burned' he was running 4.99Ghz at 1.599v.....and 5Ghz+ at 1.64v????
http://outofspecs.gr/forum/showpost....6&postcount=23

[Neuromancer]

Guess he works for Foxconn?

[Raul-7]

That was due to user error not a design flaw, it still looks better than the burned Foxconn sockets I've seen.

[EarthDog]

Originally Posted by Raul-7

That was due to user error not a design flaw, it still looks better than the burned Foxconn sockets I've seen.

Can you support that statement?

[freeagent]

Interesting..

I dont really see the big deal. Tho I can see it being a problem.. Ive seen products from both intel and amd, high end and low end have electrical problems over the years.. Not so much in the cpu socket area, but in other "just as critical" spots. It just so happens this type of problem perhaps is brought on by user error, rather then it being an electrical design fault..

[I.M.O.G.]

Updated thread title. Please in the future, try to use thread titles which are reasonable.

[hokiealumnus]

It could also be the apparently reduced number of current-carrying pins. Fewer pins + similar power requirements (at very high overclocks) could = cinged electrodes, very easily.

Seriously guys...user error? It's pretty hard to put one of these CPUs in incorrectly. I mean REALLY hard. You'd have to either break the CPU PCB or bend the loadplate if it were in wrong. Inserting these things is not rocket science. It may not be a flaw with the socket so much as too few pins to transfer the amount of current, but it's very doubtful this is user error. It also can't really be considered too few pins...at stock this would never happen. Only when running out of spec does such a thing occur.

[chewonthis]

I really think this has more to do with amps draw than anything.

[Neuromancer]

Originally Posted by chewonthis

I really think this has more to do with amps draw than anything.

QFT

Lower voltage requirements and same wattage requirements = BOOM

(yes I know materials have advnaced, and the wattage reference was to previous gen not socket 1366 LOL)

[freeagent]

I see what you guys are saying, I was just thinking maybe 1 side was clamped down more then another causeing the cpu to rock in its socket.. I did not realise the requirements are as similar as they are, I should have known better lol

I need some food.

[Raul-7]

Originally Posted by EarthDog

Can you support that statement?

Sorry for the poor choice of words. I meant in the sense that he went over the safety voltage limit set by Intel. I'm still a noob, but I thought it is recommended to keep the voltage under 1.45V on a LGA 1156 socket?

[gwar2288]

Originally Posted by Raul-7

Sorry for the poor choice of words. I meant in the sense that he went over the safety voltage limit set by Intel. I'm still a noob, but I thought it is recommended to keep the voltage under 1.45V on a LGA 1156 socket?

Yes. And you know when Foxconn Sockets exploded? When people were pumping large voltages through them!

[Daddyjaxx]

Sorry...I have no empathy. If you buy a CPU designed to run at 2.66, if it runs at 2.66 with stability, who can you cry to? You can't go back and scream "defect" when you are running massive amounts of voltage trying to push the hardware. You overclock...you roll the dice. Your CPU frys, you buy another. I really doubt that Intel's safety zone is designed for overclockers in mind. Tell Intel that you ran 1.6v through an I5 and I guarantee the warranty reject button will be hit faster than you can fathom.

[wickedout]

Originally Posted by gwar2288

Yes. And you know when Foxconn Sockets exploded? When people were pumping large voltages through them!

If that's the case then extreme overclockers will lose out and have no CPU's to test.

Intel states the limits for their CPU voltages!

[gwar2288]

Originally Posted by Daddyjaxx

Sorry...I have no empathy. If you buy a CPU designed to run at 2.66, if it runs at 2.66 with stability, who can you cry to? You can't go back and scream "defect" when you are running massive amounts of voltage trying to push the hardware. You overclock...you roll the dice. Your CPU frys, you buy another. I really doubt that Intel's safety zone is designed for overclockers in mind. Tell Intel that you ran 1.6v through an I5 and I guarantee the warranty reject button will be hit faster than you can fathom.

Oh I agree, I am just sick of all the "omg Foxconn are t3h ghey" that has been crawling about the interwebs lately

[FlahsMemory]

High over clocking is not the only way these sockets are getting blown up or burnt.There is an article I read over at THG. If I remember correctly, it stated that at voltage less than 1.4v they noticed areas turning black and burning after 1000 hours of use. In the article there is mention of some pins not aligning properly to the grid. Others reviewers say it was the foxconn socket, others say it is because of the reduce pin count etc. All in all I wouldnt consider buying the 1156 today. There is just too much problem with it and a year from now I wouldn't want to risk malfunction and frying other system components due to their mistakes.

Supposedly Intel is working on a newer revision B and C of this socket.

[EarthDog]

@ gwar - But that was the issue initially. It was only Foxconn sockets that were publicly having the issue until you saw that other thread and posted this one here. Like Anand's article regarding the Foxconn socket stated, this will likely only affect the extreme (meaning post water) overclockers with extreme type voltages. Now we have one public posting of it happening to a Lotes socket.

Have there been any public threads about this happening with air/water overclocks?

I have been having a great discussing with Chew on this via PM and now I must admit that I am leaning towards his side of the tracks in not even reccomending this socket until at least a lot more time goes by to see if normal overclocks start cooking the sockets. Mad respect for Chew, and Im not trying to minimize his opinion at all, but I would like a larger cross section of 'expert' opinions on this matter. It would help in getting an even more accurate big picture of the issue.

[chewonthis]

Originally Posted by gwar2288

Oh I agree, I am just sick of all the "omg Foxconn are t3h ghey" that has been crawling about the interwebs lately

Heh I never blamed foxcon from the getgo, I never blamed lotes either. I think the problem has always lied with the 1156 power scheme.

Foxconn sockets seemed to get trashed rather hard when it happens, the "unreported" info i have seen on lotes has been like the above picture deformed or blackened cpu pads and slightly discolored contact pins but no sockets melting.

Technically you can't blame intel from a stock consumer stand point but you can for marketing something as a budget ocing setup.

The real question is going to be if these are so short fused in the extreme ocing department, whats going to happen to average joes a year from now or even the air/water guys with moderate OC's?

1366 and 920 DO is still the best bang for buck intel setup I will reccomend.

Bottom line it's proven.


EDIT- I do find it rather odd the way that particular socket above burned....more importantly where it burned.

[wickedout]

Originally Posted by chewonthis

Heh I never blamed foxcon from the getgo, I never blamed lotes either. I think the problem has always lied with the 1156 power scheme.

Technically you can't blame intel from a stock consumer stand point but you can for marketing something as a budget ocing setup.

The real question is going to be if these are so short fused in the extreme ocing department, whats doing to happen to average joes a year from now or even the air/water guys with moderate OC's?

1366 and 920 DO is still the best bang for buck intel setup I will reccomend.

Bottom line it's proven.

The average Joe could run with a 1156/i series set up and never run into any problems. Why hasn't any of the manufacturers done a recall?

It's only HAPPENING with EXTREME OVERCLOCKERS going beyond Intel's SPECS for this CPU! How hard is it for us to understand that fact?

Also as far as I'm concerned my rig is holding up well with it's Lotes sockets!

[Daddyjaxx]

The problem is that for the "unknowledged" I would guarantee that most of the general public thinks that 1156 is the replacement for 1366. It's newer..so it's the replacement and better...right? Marketing 1156, from what I have seen, is pushing that legend too.

You want to be an early adopter, you get the pleasures of finding the issues.

[chewonthis]

Originally Posted by wickedout

The average Joe could run with a 1156/i series set up and never run into any problems. Why hasn't any of the manufacturers done a recall?

It's only HAPPENING with EXTREME OVERCLOCKERS going beyond Intel's SPECS for this CPU! How hard is it for us to understand that fact?

Also as far as I'm concerned my rig is holding up well with it's Lotes sockets!

Ok you seem to be misinformed.......You do realize Intel and AMD both use overlockers for various reasons.......one of them is to find the breaking points.

Reason being they may find out issues that could happen years down the line due to prolonged use at just stock.

[wickedout]

Originally Posted by chewonthis

Ok you seem to be misinformed.......You do realize Intel and AMD both use overlockers for various reasons.......one of them is to find the breaking points.

Reason being they may find out issues that could happen years down the line due to prolonged use at just stock.

Answer me this. Why hasn't there been a recall?

How many actually do overclocking anyways? The i5/i7 will not be recalled because the average user doesn't overclock. We are talking masses here.

Do you see my point?

[Grnfinger]

Originally Posted by wickedout

The average Joe could run with a 1156/i series set up and never run into any problems. Why hasn't any of the manufacturers done a recall?

It's only HAPPENING with EXTREME OVERCLOCKERS going beyond Intel's SPECS for this CPU! How hard is it for us to understand that fact?

Also as far as I'm concerned my rig is holding up well with it's Lotes sockets!

I couldnt agree more, ppl tend to exaggerate things on the internet.
My foxconn socket is working 100% I have checked it 3 times since I bought the board but then again I dont run 1.65 volts to my cpu...

I keep hearing the same thing over and over 1156 oh dont get a foxconn socket! and its usually from ppl who have never even tested the platform ( not directed towards anyone here) ...imo if it were wide spread and a serious issue the major manufacturers would have issued a recall

I think Chew has the best argument so far

[Raul-7]

Originally Posted by wickedout

The average Joe could run with a 1156/i series set up and never run into any problems. Why hasn't any of the manufacturers done a recall?

It's only HAPPENING with EXTREME OVERCLOCKERS going beyond Intel's SPECS for this CPU! How hard is it for us to understand that fact?

Also as far as I'm concerned my rig is holding up well with it's Lotes sockets!

So far it's only happened to one LOTES socket and he was pushing it way beyond the breaking point. But still, the burn is nowhere near as bad as the burns I've seen on the Foxconn sockets.

OT, Good to see MSI switching over to LOTES. I wish ASUS would do likewise.

[Grnfinger]

Originally Posted by Raul-7

So far it's only happened to one LOTES socket and he was pushing it way beyond the breaking point. But still, the burn is nowhere near as bad as the burns I've seen on the Foxconn sockets.

OT, Good to see MSI switching over to LOTES. I wish ASUS would do likewise.

One reported case so far and I dont see the internet littered with foxconn issues, just the rumour getting played up.
I will sleep easy knowing I have a 3 year warranty and not likely to have the board a year

[Neuromancer]

Originally Posted by Grnfinger

I couldnt agree more, ppl tend to exaggerate things on the internet.
My foxconn socket is working 100% I have checked it 3 times since I bought the board but then again I dont run 1.65 volts to my cpu...

I keep hearing the same thing over and over 1156 oh dont get a foxconn socket! and its usually from ppl who have never even tested the platform ( not directed towards anyone here) ...imo if it were wide spread and a serious issue the major manufacturers would have issued a recall

I think Chew has the best argument so far

The point is not how prolific is it, the point is it happens. Would you rather no one ever bring it up? It is naive to think that by not knowing about it it will never be an issue.

I have not seen anyone say, OMG dont get a P55 because they blow up!! Seen lots of posts saying, if you are getting a P55 try and get one with a LOTEs socket.

Also sockets are not hte only thing to blow up on P55 either. Even a build overvolting of 1.4volts was enough to pop a FET on the Biostar P55 just like happened to some people on the AMD boards that were supposedly 140W compatible.

It lets you know... hey, going to OC make sure your FETs are sinked (obvious to some not to all) same with the possibility of OCing on P55, you are far more likely to over amp the socket wit h a foxconn built, over a LOTES.

Sorry, why would having that information out there be bad? You think Foxconn is going to go out of business over it? If they went under prices would soar from the vacuum they made

[chewonthis]

Originally Posted by wickedout

Answer me this. Why hasn't there been a recall?

How many actually do overclocking anyways? The i5/i7 will not be recalled because the average user doesn't overclock. We are talking masses here.

Do you see my point?

Ok well first of all a recall is up to any manufactures own discretion if not a safety recall........

I have seen no cases of peoples houses burning down.......

Secondly there has to be a certain amount of instances where peoples safety were at risk to force a recall.

The fact that intel is working on a revision is really the proof in the pudding.

It's a problem that can get sidestepped, they have limited cases of it on rev A revise it with Rev B and it never happens again.

For the record here is my foxxconn socket On the "guilty" 1156 board.

YES believe it or not I have a UD6 and an 860.



And here is a rather high leakage ES 860.



Both of which have shown no problems thus far.

The question is why and that is really what is trying to be determined.

You can't fix something if it can not be properly diagnosed.

[Raul-7]

Any more facts on this revision? Is it for the socket only or other components? ETA? etc.

[chewonthis]

Originally Posted by Raul-7

Any more facts on this revision? Is it for the socket only or other components? ETA? etc.

I really have no clue what they are revising, I have just heard that there will be one.