[SOLVED] Rust/corrosion in CPU pins. how to handle it?

[Sarathydn]

I did not use my PC for nearly 4 months. after that(now) i tried to turn my PC, it is running, but the screen is blank. not even bios screen.
tried with different cable, monitor. then i cleaned the RAM and changed battery.

same problem. then i took it a technician and found there is corrosion in few of my CPU pins. then i searched internet and came across some thread in tom's hardware too.

then i bought IPA cleaning reagent to clean the CPU Pins.

Now,
1) I need some suggestion/guidance on cleaning the rust/corrosion. shall i use something like a blade or tips of forceps to gently scratch away the rust marks?

2) after socking only the tips of the CPU in IPA, is it enough to brush the pins with new tooth brush to remove the rust or is should use brass brush ?

3) as the pins having rust, would the socket also have residues of rusts in it? if yes, can it be removed?

3)is it possible to change the socket if needed ?

4) will my PC be alright after cleaning that corrosion? or is there is anything i have to do in order to prevent issue in the future

Motherboard : Asus M5A88-V EVO

Processor : AMD-FX 8-Core Processor (Black Edition) FX8150

 

[Grobe]

That sound very strange - I've never heard of that issue before. Assuming the technician is correct, I'd say AMD have a far greater problem than you, assumed you've storing the computer indoor and in a proper environment.

Also - if you're taking on trying to brush it clean, there is a great chance of just destroying it.

 

[Gam3r01]

Can you post a picture of said corrosion? That sounds very odd.

 

[Karadjgne]

The pins are brass. Brass doesn't rust or corrode like that. And certainly shouldn't suffer any such damage in just 4 months.

Sounds more like a capacitor on the motherboard has leaked/failed and what you are looking at is the leakage results that have trickled/siphoned into the socket.

Meaning the motherboard is basically toast.

 

[Name doesnt matter]

The pins are brass. Brass doesn't rust or corrode like that. And certainly shouldn't suffer any such damage in just 4 months.

Sounds more like a capacitor on the motherboard has leaked/failed and what you are looking at is the leakage results that have trickled/siphoned into the socket.

Meaning the motherboard is basically toast.

But how it is possible capacitors usually start dieing after 10 years.
Edit :Not 10 years lets say after 15 years assuming you have worst capacitor in you mobo

 

[drea.drechsler]

The pins are brass. Brass doesn't rust or corrode like that. A
..

Slight clarification: the pins are gold plated brass. Gold is a noble metal, it absolutely won't corrode or 'rust', and is used there specifically for that reason. Brass won't 'rust' either, of course, but it will oxidize.

I'd also want a picture to know what's going on; it's possible the plating has failed and is peeling off the pins leaving exposed brass to tarnish (oxidize) and look like corrosion. Same effect, of course.

But how it is possible capacitors usually start dieing after 10 years.
Edit :Not 10 years lets say after 15 years assuming you have worst capacitor in you mobo

The caps that die in 10/15 years are the electrolytic caps used in VRM filtering. They pass heavy current in that application, and the dielectrics will 'punch through' with time, temperature and voltage. Also, seals used in wet electrolytics will fail and let the electrolyte leak out. Better/more expensive caps have higher voltage tolerance and longer life.

 

[Karadjgne]

Slight clarification: the pins are gold plated brass. Gold is a noble metal, it absolutely won't corrode or 'rust', and is used there specifically for that reason. Brass won't 'rust' either, of course, but it will oxidize.

That's what I was thinking. It's an FX, and it's not pristine, and the socket uses pin clamps to hold down the cpu, possibly scratching off the plating which is only a few molecules thick to start with.

But how it is possible capacitors usually start dieing after 10 years.
Edit :Not 10 years lets say after 15 years assuming you have worst capacitor in you mobo

Caps can die at any time. It's an electronic component. Brand new motherboards fail at first turn on because a cap has failed. 15 years of electrical abuse is a lot when you consider most psus are only now warranted for 5+ years, back when that mobo was new you'd be looking at good psus with a 1 or 3 year warranty, just because of capacitor degradation and failure. Same caps on a motherboard are found in psus, although a psu also has some larger ones.

You can always tell when a motherboard is about to fail, look at the caps. If they are no longer 'tube' shaped, but have swollen to look like a 'barrel' instead, they are dying inside. If the X on the top is splitting, they are dying.

Your cpu is 125w. Its using 1.4(ish) volts to get that 125w. W=VxA, so 125=1.4xX or X=125/1.4 =90A.

Your household electric oven uses 40A. Most air-conditioning systems on heat settings use less than 80A. You've got a motherboard with a tiny cpu that's demanding 90A.

Caps fail. They leak. They die. And your cpu is right below a very large bank of VRM caps. Leakage into the socket after 15 years is very possible.

 

[drea.drechsler]

...
possibly scratching off the plating which is only a few molecules thick to start with.
....

I've never seen a good spec on CPU's to know...but usually commercial electronics is somewhere in the 10- 30mil thickness range. A bit more than a few molecules but still pretty darn thin. And that's over a nickel layer as gold doesn't work well plated directly on brass.

 

[Karadjgne]

Gold plating is the deposit of a thin layer of gold metal on a substrate, typically in the 0.25 – 5 micron range (0.00001”-0.0002”).

By comparison, a human hair is roughly 0.001574803" thick or roughly 10x to 150x larger than the gold plating on a cpu pin.

Couple molecules thick, that's it. The nickle won't be much thicker. Gives an entirely new perspective on the term 'paper thin'. (0.0039")

 

[drea.drechsler]

Gold plating is the deposit of a thin layer of gold metal on a substrate, typically in the 0.25 – 5 micron range (0.00001”-0.0002”).
...

hmmm...cpu's much be crappy bad then.

my experience was

Gold plating is the deposit of a thin layer of gold metal on a substrate, typically in the 0.25 – 5 micron range (0.00001”-0.0002”).

By comparison, a human hair is roughly 0.001574803" thick or roughly 10x to 150x larger than the gold plating on a cpu pin.

Couple molecules thick, that's it. The nickle won't be much thicker. Gives an entirely new perspective on the term 'paper thin'. (0.0039")

A gold atom is .2882 nanometer diameter.... or 0.0000000113465". I make that about 881 atoms thick if it really is .25 micron thick, which is the lowest spec I've ever seen on any contact. I really suspect it's closer to 10 mil thick, maybe as high as 20 mil. That's in the 'thin' gold plating range for commercial products.

At any rate 881 atoms is a bit more than a few, even then.

 

[Karadjgne]

Yeah lol, it's pretty thin, especially considering the size of the individual pins themselves. The brass/nickle making up all the strength, if the gold were too thickly layered they'd have to use even thinner diameter base pins to allow for the socket entrance.

For a connectors and contacts with moderate environmental and wear cycles, common functional gold plating thickness ranges between 30-50uin (0.75-1.25um).

1um = 1 micron. So by that token, the middling plating is @ 1 micron avg. The thinner stuff being nothing more than gold polka-dot, the thicker plate too soft to support the pins rigidity. It's the little dark lines you see on the microscope views of hair on CSI. If that thick lol.

 

[Sarathydn]

Can you post a picture of said corrosion? That sounds very odd.

thank you very much for everybody's help in this issue

Pics of CPU Pins

as there is no option to insert image directly from computer, i am inserting link from Google Photos. if it doesn't work, pls tell me how to insert a pic.

and i have an update to...

i tried gently to remove rust using a pen knife, and i hope i done it right.. posting a pic of that too..anybody kindly enlighten me if i have to do anything else.

Pics after cleaning with IPA

 

[Gam3r01]

Well, that is certainly interesting. Ive personally never seen a processor corrode like that in a typical environment (wet storage locker, yeah sure, but this?)
You did well attempting to remove it, I dont see anything to indicate any bent or broken pins, so good job on that.
The markings on the side of the IHS are what concern me, that is a fairly substantial buildup with no explanation as to where it came from.

That being said, that corrosion did not look severe enough to be causing issues. I worry that the socket itself is where most of the issue is. At this point the best you can do is try it out and see what happens.

 

[Sarathydn]

thank you for acknowledging my attempt... the dirt like markings on the side of IHS is simply dust particles only.. i ve cleaned it after u mentioned it.

i stored in my PC in normal room temperature. that is actually living room.

and before trying the CPU , i doubt the socket that it may have rust residues . if am right, is there any way to remove that?
For Example : spraying IPA into the socket? or using absorb mode in vaccum to remove loose rust particles??

 

[Sarathydn]

Slight clarification: the pins are gold plated brass. Gold is a noble metal, it absolutely won't corrode or 'rust', and is used there specifically for that reason. Brass won't 'rust' either, of course, but it will oxidize.

I'd also want a picture to know what's going on; it's possible the plating has failed and is peeling off the pins leaving exposed brass to tarnish (oxidize) and look like corrosion. Same effect, of course.



The caps that die in 10/15 years are the electrolytic caps used in VRM filtering. They pass heavy current in that application, and the dielectrics will 'punch through' with time, temperature and voltage. Also, seals used in wet electrolytics will fail and let the electrolyte leak out. Better/more expensive caps have higher voltage tolerance and longer life.

Pics of CPU Pins

and i have an update too...

i tried gently to remove rust using a pen knife, and i hope i done it right.. posting a pic of that too..anybody kindly enlighten me if i have to do anything else.

Pics after cleaning with IPA

 

[SkyNetRising]

Pics of CPU Pins
and i have an update too...
i tried gently to remove rust using a pen knife, and i hope i done it right.. posting a pic of that too..anybody kindly enlighten me if i have to do anything else.
Pics after cleaning with IPA

You have serious issue with humidity levels inside your pc.
What exactly is your housing situation? Flooded basement or something?

 

[SamirD]

That's a lot of corrosion and it could be a number of things that are wrong at this point. You've cleaned up the pins pretty well and as long as the socket doesn't have any hidden corrosion you should be good to go. One thing I would definitely do is socket and unsocket the processor several times before trying to boot as it will help with any potential corrosion inside the socket.

I would also check all other metal in this system for corrosion and check for a hidden water leak in your house.

 

[tafsirnahian5]

I can assure that rust is preventing your computer to turn on. I've similar problem with my AMD sempron 145 processor. I have two processor. One is AMD phenom X2 545 and one is sempron 145. The sempron has developed rust on its pin. At least 30 pins are affected with rust. So with the sempron CPU, the system doesn't turn on. No display signal. But if I plug the phenom the system works just fine.