Intel LGA 1700 socket problem??

[elgato]

Yes, horrible picture, sorry. Quick pic with an Iphone 6 before boxing it up.
There was no damage when the CPU was installed. Frightening that just lifting the CPU out could damage the socket. Its like the CPU pins bonded to the socket traces and pulled them up.
Just a defect? One time failure?
MSI socket issue that has not been widely reported yet?
LGA 1700 socket issue that has not been widely reported yet?

 

[ir_cow]

I haven't damaged a socket besides dropping the CPU on it. Bent pins is user error generally.

 

[elgato]

Notice how they are pulled up? I still think that either they stuck to the processor and were lifted up when it was removed or they
were loose to begin with. This socket is a whole new design. Also, if they were damaged as bad as they look there is no way it would have ran perfectly for two weeks.
It was removed by simply unlocking the latch and carefully lifting it straight up.

 

[Totally]

I haven't damaged a socket besides dropping the CPU on it. Bent pins is user error generally.

Yeah but when have the other sockets suffered damage from dropping a cpu on it? I've dropped cpu on sockets pga sockets more than I can remember and never damaged the socket nor cpu. Even when a cpu fell off the side of a desk once resulting in a row of bent pins, the issue was easily resolved with click pen. So that aspect of the less resilient design can't be written off a user error.

 

[Deleted member 24505]

The socket certainly is not a new design. I can't think of any way the pins in the socket would bond to the CPU contacts unless they were power ones and somehow overloaded and "welded" themselves to the CPU. I would have liked a better pic.

My guess is, on return, they will say "user error"

 

[Totally]

The socket certainly is not a new design. I can't think of any way the pins in the socket would bond to the CPU contacts unless they were power ones and somehow overloaded and "welded" themselves to the CPU. I would have liked a better pic.

My guess is, on return, they will say "user error"

They could have cold welded if the the points of contact were clean enough, unlikely but plausible

 

[elghinnarisa]

They could have cold welded if the the points of contact were clean enough, unlikely but plausible

It sure as heck couldn't. Unless they kept their computer in a vacuum, which i highly doubt.

 

[Totally]

It sure as heck couldn't. Unless they kept their computer in a vacuum, which i highly doubt.

Presence of vacuum isn't required for a cold weld to occur. Being in vacuum just reduces the probability of a foreign atom coming in between the two metal atoms or reacting with the metal atoms to zero.
Also gold does not form an oxide layer.

Cold Welding in Atmosphere (vacuum tests to come)

 

[Mr.Scott]

Irrelevant how it happened. Every manufacturer out there will reject RMA for damaged socket pins. User error every time.

 

[Mussels]

The socket certainly is not a new design. I can't think of any way the pins in the socket would bond to the CPU contacts unless they were power ones and somehow overloaded and "welded" themselves to the CPU. I would have liked a better pic.

My guess is, on return, they will say "user error"

12th gen does use a new socket design, which can actually bend the CPU's themselves (see igors lab)


Any previous socket, i'd be saying user error... but with a socket that flexes and bends CPUs? Might be a design flaw.

 

[Deleted member 24505]

12th gen does use a new socket design, which can actually bend the CPU's themselves (see igors lab)


Any previous socket, i'd be saying user error... but with a socket that flexes and bends CPUs? Might be a design flaw.

I meant as in the LGA socket. It might be slightly different, but it is not really a new design as it is still LGA

From what i read, it was cheaper boards that was the problem, not the socket. As far as i know there hasn't been any high end board with socket problems has there? Wasn't it a cheaper board that had the problem. Otherwise why would every LGA 1700 socket suffer the same? I have a really thick EK backplate on my board, and my chip is still perfectly flat.

 

[Totally]

I meant as in the LGA socket. It might be slightly different, but it is not really a new design as it is still LGA

From what i read, it was cheaper boards that was the problem, not the socket. As far as i know there hasn't been any high end board with socket problems has there? Wasn't it a cheaper board that had the problem. Otherwise why would every LGA 1700 socket suffer the same? I have a really thick EK backplate on my board, and my chip is still perfectly flat.

In the write up he mentioned that the more expensive boards are affected it's that the the boughing is less pronounced and needs to test further if their adversely affected and if so by how much. He made it clear that it was not the cheap boards that are the problem, it's the high clamping pressure of LGA1700. Just merely dropping the cpu, closing the retention clip and presto a bendy board, some more bendy than others. It's all there clearly written how did you miss all that? What about folks using stock coolers or air coolers that don't utilize a backplate? Fk 'em bc your wb has one thus sidesteps the issue?

 

[Deleted member 24505]

Fk 'em bc your wb has one thus sidesteps the issue?

Did i say that NO i never. What makes you think i don't give a fuck as long as i am ok. I stated mine did not have a problem, maybe because of the thick backplate. I was wondering why everyone on TPU with a LGA 1700 setup who hasn't got a water block on their system hasn't had the problem happen. Jesus some fucking people i think maybe you need another coffee or something

It was because of this-If you take a rather “cheap” Z690 board, in this case an ASRock Z690 Extreme, Incidentally, system integrators also pointed out that the rather cheaper boards with thinner PCBs already had a U-shaped bulge around the socket before a CPU was even inserted. that i said wasn't it a cheaper board that had the problem. I have just read through the page i originally linked and he never says more expensive boards have the same problem. Maybe you need to go back and re read it too.

 

[RJARRRPCGP]

I was having issues with my MSI Z690 Edge Wifi DDR4, I7 system locking up.
After a lot of diagnosis the Motherboard became suspect.

This afternoon I disassembled the board for return and found this damage when I removed the CPU.

In my experience, once it's in the socket, I never had a failure of socket pins. In fact, in that photo, it looks just like pre-CPU-installation damage!

In fact, a properly-installed CPU, acts like a shield, with Intel LGA! This is why I like getting second-hand Intel-platform motherboards with a CPU installed in it!
For example, getting a socket 1366 motherboard with a Core i7 920 in it.

Irrelevant how it happened. Every manufacturer out there will reject RMA for damaged socket pins. User error every time.

Except for January, 2013, when I had to have the motherboard returned for repair at the least, because the Asus Maximus II Gene, was advertised as working. And, I did get a properly working one on February 11, 2013. It took weeks, like it was sent to China, but getting a working one a month later, is a ton better than never!

I narrowly escaped the worst that I feared. I was very close to having to keep my Asus P5QL Pro. (P43)

Today, 9 years ago, there were unusually high temps outside (52 F! A record for Springfield, Vermont on January 14, 2013) and I went for a walk, to come back to the motherboard always failing to boot, I got nothing but fan and LEDs! Then, I saw the terrible-looking socket!

 

[abrusco]

I had a problem with my motherboard after putting in the CPU. In addition to the bent pin, I also had problems with one of the cooler pins that didn't fit.

Board B660M DS3H DDR4 (rev. 1.0)
CPU I5 12400
Sorry for my english

 

[ThrashZone]

Hi,
I never liked the crappy sockets on intel boards
Only intel hedt x99/ x299 boards seem to be best designed sockets.

 

[Deleted member 24505]

Hi,
I never liked the crappy sockets on intel boards
Only intel hedt x99/ x299 boards seem to be best designed sockets.

Never really had a problem with them myself. What's better about the X99 and 299 Intel sockets apart from the size and the contradiction.

 

[ThrashZone]

Never really had a problem with them myself. What's better about the X99 and 299 Intel sockets apart from the size and the contradiction.

Hi,
Both x99 and x299 have built in back and front plates z series have nothing.
x299 on left z490 on right

 

[Totally]

Again remind me again why lga is better than pga?

 

[ThrashZone]

Hi,
The four cpu mounting holes are pretty obvious which one is stronger and less likely warp the board
x99/ x299 four holes are in the metal plate
Z series are just four holes through the thin board
On each board the four holes are the same distance apart

 

[darkbreeze]

I think you misread my comment. Try again and you might get it.
Lotes seem to be the main socket supplier at the moment, with Helm Technology being the other one, as in the pictures above.

Actually, according to Igor and Buildzoid, the two socket suppliers are Lotes and Foxconn, not Helm.

I meant as in the LGA socket.

What you meant and what you said are two different things. What you SAID was it is not a new socket, however it 100% IS a new socket. What is not new is the array type, since it is still a Land grid array.

The socket, socket 1700, is totally new and has not ever been used for any previous CPU design or generation. Which means there can certainly be a number of growing pains especially since this specific socket has already proven to have some fundamental engineering and design issues and a higher pin density than previous generations, by 500 pins.

 

[Deleted member 24505]

Actually, according to Igor and Buildzoid, the two socket suppliers are Lotes and Foxconn, not Helm.


What you meant and what you said are two different things. What you SAID was it is not a new socket, however it 100% IS a new socket. What is not new is the array type, since it is still a Land grid array.

The socket, socket 1700, is totally new and has not ever been used for any previous CPU design or generation. Which means there can certainly be a number of growing pains especially since this specific socket has already proven to have some fundamental engineering and design issues and a higher pin density than previous generations, by 500 pins.

LGA is not new, whether it has more pins or not, still LGA.

 

[darkbreeze]

I was only clarifying that one of the stated socket suppliers mentioned by TheLostSwede was inaccurate from what they actually were according to two pretty distinctly accurate sources, and had nothing whatsoever to do with you AND then, in regard to what YOU said, specifically, was:

The socket certainly is not a new design.

And the fact is, it IS, and was merely clarifying on that account as well. I specifically stated that Land grid array, which is an array, not a socket, isn't new, but that Socket 1700 IS new. That's it. That's ALL I said.

I didn't attack or in any way disparage you personally, I simply clarified some accidental misinformation. And I certainly, 100% did not say that LGA was new. I specifically stated that it isn't, because Land grid array has been around for a very long time. There is definitely not a need for you to be getting personal. It's uncalled for.

And further, people coming along to upvote posts that are personal attacks, is also, or at least should be, uncalled for and sanctionable. I've worked as a moderator on another very high profile tech forum for a really long time, and we sure wouldn't tolerate those kinds of behaviors, especially out of the blue for no reason.

 

[Deleted member 24505]

I was only clarifying that one of the stated socket suppliers mentioned by TheLostSwede was inaccurate from what they actually were according to two pretty distinctly accurate sources, and had nothing whatsoever to do with you AND then, in regard to what YOU said, specifically, was:

And the fact is, it IS, and was merely clarifying on that account as well. I specifically stated that Land grid array, which is an array, not a socket, isn't new, but that Socket 1700 IS new. That's it. That's ALL I said.

I didn't attack or in any way disparage you personally, I simply clarified some accidental misinformation. And I certainly, 100% did not say that LGA was new. I specifically stated that it isn't, because Land grid array has been around for a very long time. There is definitely not a need for you to be getting personal. It's uncalled for.

And further, people coming along to upvote posts that are personal attacks, is also, or at least should be, uncalled for and sanctionable. I've worked as a moderator on another very high profile tech forum for a really long time, and we sure wouldn't tolerate those kinds of behaviors, especially out of the blue for no reason.

Just because they increase the pin count and it is called 1700 it is still LGA so not new, Maybe we will have to agree to disagree.

 

[eidairaman1]

Hi,
I never liked the crappy sockets on intel boards
Only intel hedt x99/ x299 boards seem to be best designed sockets.

The TR method is pretty fool proof.