Sign in with
Sign up | Sign in
Your question

Broke phenom II

Tags:
  • CPUs
  • Phenom
Last response: in CPUs
Share
January 8, 2009 8:12:35 PM

I saw them today and Decided to go for the phenom II but when i installed it on my mobo 2 pins broke off... are they faulty?

More about : broke phenom

January 8, 2009 8:28:49 PM

First off - Phenom II was only just released today - and I highly doubt you bought it out of a shop . So either (I'm not saying you are) you're lying or you've mistaken a Phenom I for a Phenom II.

What ever the case pins don't break unless you're a complete numpty and make them break - unless they fell off the die perfectly. Which I doubt - unless you've a screenie to prove otherwise. So the fact is, the chip isn't faulty - you're just heavy-handed.
January 8, 2009 8:53:58 PM

I'd agree with Roffey123, hard to bend cpu pins unless you're careless.

Whatever cpu it was, it's just a square of junk now, I doubt you'd get a refund......
Related resources
a b à CPUs
January 8, 2009 9:01:04 PM

TROLL.
January 8, 2009 9:21:03 PM

LOL, whats worse... the pins breaking or telling people you broke it?

RMA that thing... tell them you thought it was Intel!
January 8, 2009 9:28:40 PM

RMA wont be accepted. Newegg or any online retailer do not refund/replace for bent/broken pins. Processors have a totally different return policy than all other products.

If this person actually does own this chip, odds are anyways they got it in a brick and mortar. Even then a return might be hard/impossible considering the box is open.
January 8, 2009 9:44:45 PM

I broke a 486 when they had just been released - in the days before zif sockets - used a screwdriver to get it out of the socket and broke a corner of the chip right off...

**** happens...
January 8, 2009 11:08:43 PM

jamesgoddard said:
I broke a 486 when they had just been released - in the days before zif sockets - used a screwdriver to get it out of the socket and broke a corner of the chip right off...

**** happens...



LOL, I misidentified pin 1 on my first genuine 486 build (not the 486DLC that I managed to do right in the prior generation). Well, I didn't bend the pins or break anything but the 486DX2 and the motherboard died and I had to spend the cash to replace them as it was my only PC at the time.

I was so happy when ZIF sockets came out. It was easier in the 386SX days when the CPU's were soldered on the motherboards and easier with ZIF. You just had to be very careful with LIF sockets.

Still, I can't see how he got a Phenom II in his klutzy hands this early. He may be making the whole thing up.

January 8, 2009 11:10:47 PM

he is just as clumsy whether it was a p1, p2 or an old Thunderbird lol
January 8, 2009 11:14:33 PM

the hammer slipped when i was installing a p4, broke the processor and the motherboard
January 8, 2009 11:18:00 PM

sometimes its better to squeeze them in with channel locks than the hammer. hammer is more of a finesse tool if the channel locks don't force the pins where they need to go.
January 8, 2009 11:24:51 PM

wow, if this is not bs, then that really sucks!!! best of luck RMA that.
a c 95 à CPUs
January 8, 2009 11:30:25 PM

A G clamp is always helpful to install cpu's that resist .

Put the cpu over the socket and then put the clamp around the mb and cpu . Tighten them up . Really tight . Its guaranteed to force the pins in properly .
This method is so versatile you can even get intel cpu's into AMD boards for extra performance .
EXTRA performance
January 8, 2009 11:34:02 PM

now thats funny!!!
January 8, 2009 11:34:54 PM

Outlander_04 said:

This method is so versatile you can even get intel cpu's into AMD boards for extra performance .
EXTRA performance


Not since the socket 7 days. :lol:  That was the last time AMD and Intel supported the same socket. You should have had a smily at the end to show you weren't serious, unless you were seriously trying to mess with the OP's mind.
January 8, 2009 11:40:55 PM

outlander is correct. the compression from the G clamp will "right" any "wrong" there was in engineering. i would dare say a Pentium 1 will fit in a 939 board using this technique. heat will be a non-issue as there will be no heat created by the CPU ever again.
January 8, 2009 11:53:16 PM

I just had a 939 msi motherboard stop working tonight, maybe I can upgrade to the new phenom II and put my 9850be in the 939 mother board and like magic it will come back to life with outlanders method!!!
January 8, 2009 11:55:32 PM

its worth a try.
January 8, 2009 11:57:55 PM

I have lapped the 9850be so the heat issue should work out even better with this new method, I am gonna try it and I will let you guys know how it turns out!! BRB :lol:  :lol:  :lol: 
January 8, 2009 11:59:55 PM

medjohnson77 said:
I just had a 939 msi motherboard stop working tonight, maybe I can upgrade to the new phenom II and put my 9850be in the 939 mother board and like magic it will come back to life with outlanders method!!!


Reverse engineer outlander's method. Pluck the pins off the CPU and insert it in a PC Chips socket T board. Then, when it's not working, you can blame PC Chips. If you can't find a PC chips board, then an ECS will do. :na: 

Note the use of a smiley, I don't want any clueless noob taking me seriously.
January 9, 2009 12:02:21 AM

Well I don't have a pc chips or a ecs, but I do have a jetway!!!LOL.
January 9, 2009 12:10:45 AM

I wonder if the OP has trouble cooking popcorn. Just remember, ^ This Side Up^
January 9, 2009 12:11:49 AM

When a old computer tech bud of mine bought a 9550 2.4ghz new awhile back he managed to bend 2 pin's from the top left hand corner of the cpu...The cpu still ran like this which shocked us both and i only found the problem when i was reapplying past on the cpu.

He is a certified tech but age has got the best of his eyes and now he let's me do all the small things for him! I used a magnifying glass and some fine tweezer's and straightened the pins out and they went into the pin holes and everything was fine.
January 9, 2009 12:19:23 AM

xx12amanxx said:
When a old computer tech bud of mine bought a 9550 2.4ghz new awhile back he managed to bend 2 pin's from the top left hand corner of the cpu...The cpu still ran like this which shocked us both and i only found the problem when i was reapplying past on the cpu.

He is a certified tech but age has got the best of his eyes and now he let's me do all the small things for him! I used a magnifying glass and some fine tweezer's and straightened the pins out and they went into the pin holes and everything was fine.


I'm 51 1/2 years old and the one year I could not do my own builds, was when I had cataracts on both eyes. Once I got surgery for both, then it was back to doing my own work.

I still chose my components. As for eyesight, I'm legally blind in my left eye from strabismus and am very careful and use a magnifier where necessary (darned USB pins). I can't see 3D movies, but wear the glasses to be able to see it in 2D (we take our son to 3D kids movies and I chose the 2D for Beowulf over 3D).

The only reason I had surgery in my bad eye, besides the huge cataract made me look like an extra in a bad hillbilly horror movie, is that I do have good peripheral vision in that eye, everything is just blurred straight on, it's something like 20/200.

That's why I don't work as a PC Tech, but started ages ago in mainframe and then moved to batch ops. I'll still keep building my own for as many years as I can. I'm very cautious with CPU's. So far, I haven't bent a pin and only fried that one 486 due to not identifying pin 1 in the LIF days.
a c 95 à CPUs
January 9, 2009 12:21:25 AM

yipsl said:
Not since the socket 7 days. :lol:  That was the last time AMD and Intel supported the same socket. You should have had a smily at the end to show you weren't serious, unless you were seriously trying to mess with the OP's mind.



Thats nonsense! I have successfully installed slot one intel chips in socket 939 amd boards using two clamps and some 5 minute epoxy . All it needs is enough pressure and anything is possible .

January 9, 2009 12:32:45 AM

Awww, just put it in a 400 ton punch press, itll go
January 9, 2009 12:46:10 AM

Any of you guys into construction, just use one of those .22 cal. nail punchs with a cap nail at the end of it instead of a concrete nail and that might just work...
a b à CPUs
January 9, 2009 1:09:01 AM

Don't forget to Epoxy the HSF for best heat transfer.

a c 95 à CPUs
January 9, 2009 1:13:55 AM

medjohnson77 said:
Any of you guys into construction, just use one of those .22 cal. nail punchs with a cap nail at the end of it instead of a concrete nail and that might just work...


a friend of mine didnt have clamps , or a press , or even a nail gun so he got inventive and used a truck . Im serious !

A couple of wooden off cuts to support the motherboard and make a ramp was all he needed . He calculated his 3 ton truck had to be traveling at exactly 73 mph and drop 6 inches to get the required pressure .

That computer is as good today as it was then .
January 9, 2009 1:40:33 AM

yipsl said:
I'm 51 1/2 years old and the one year I could not do my own builds, was when I had cataracts on both eyes. Once I got surgery for both, then it was back to doing my own work.

I still chose my components. As for eyesight, I'm legally blind in my left eye from strabismus and am very careful and use a magnifier where necessary (darned USB pins). I can't see 3D movies, but wear the glasses to be able to see it in 2D (we take our son to 3D kids movies and I chose the 2D for Beowulf over 3D).

The only reason I had surgery in my bad eye, besides the huge cataract made me look like an extra in a bad hillbilly horror movie, is that I do have good peripheral vision in that eye, everything is just blurred straight on, it's something like 20/200.

That's why I don't work as a PC Tech, but started ages ago in mainframe and then moved to batch ops. I'll still keep building my own for as many years as I can. I'm very cautious with CPU's. So far, I haven't bent a pin and only fried that one 486 due to not identifying pin 1 in the LIF days.


It doesnt help that he has fat fingers too. My hands are big but my finger are long and skinny so i can get into area's his fingers wont fit...lol

It's funny but i bought a micro atx case thinking it would be cool saving space and being small! And i swear im going back to a mid tower becuase i cant do anything without flipping the case on it's side! Plus this thing run's super hot no matter what i do!
January 9, 2009 2:11:12 AM

speaking of breaking pins, I had a leadtek geforce2 mx that required me to break off several pins on my 19" viewsonic crt monitor in order for it to work.
January 9, 2009 5:19:26 AM

If that's so, and not a continuation of the absurd humor in this thread, wouldn't it have caused a problem when you upgraded to your next GPU?

I haven't had a Gf2mx in years, but I had no issues with several 15" and 17" CRTs. No issues with a 7600gs or with any ATI card from the Radeon 64 DDR to the 3870x2.

The way things are going here, we should offer an "Elmer Fud" award for the biggest and most outrageous lie that has a modicum of believability.
January 9, 2009 1:01:35 PM

no, I later upgraded to a geforce 4 that eventually crapped out and then later to a geforce 6800gt and it worked fine.
January 9, 2009 1:39:07 PM

I have a P4 1.8GHz chip around here that has a broken pin, I mean completely missing. It ran without problem with that broken pin for years. I replaced it with a 2.6ghz P4, but the broken pin proc is still around here... somewhere...
January 9, 2009 2:50:39 PM

940 pins on a Phenom 2 I guaruntee you that 70% of those pins are used for testing. They are used in manufacture to make sure it works. They are also used by the bios of the mobo to check things are in order. One pin is cabanle of doing many different things.

You'll be surprised how many you can break and these things will still work. I would suggest the OP tries pulling them all out 1 at a time and let us no just how many it takes. Smiley face not necassary.
a b à CPUs
January 9, 2009 3:03:12 PM

gpippas said:
940 pins on a Phenom 2 I guaruntee you that 70% of those pins are used for testing. They are used in manufacture to make sure it works. They are also used by the bios of the mobo to check things are in order. One pin is cabanle of doing many different things.

You'll be surprised how many you can break and these things will still work. I would suggest the OP tries pulling them all out 1 at a time and let us no just how many it takes. Smiley face not necassary.


uh.... lol. has this ever been confirmed or is it more trolling?
January 9, 2009 6:08:45 PM

werxen said:
uh.... lol. has this ever been confirmed or is it more trolling?


Telling the OP to pull the pins out was a joke.

What I was actually saying about the pins being used for testing however isn't. I currently have the task designing and simulating a 4 bit ALU. So incredably basic compared to a modern cpu. It has 5 arithmatic functions and 4 logic functions. The design is very easy. However I have to write about 15 test vectors for each function so therefore 135 test vectors. All the functions can be implemented using 9 pins of the 24. The rest are used for for the test vectors. So are therefore not necassary if you like. My guess is that AMD and INtel use an advanced form of scan path testing.

In basic form:

1. Place the circuit in test mode and fill the registers with a set of data.
2. Place the circuit in operational mode for 1 clock cycle.
3. Place cicuit back into tect mode and clock the data out through the scan path.
4. Repeat as necassary.

From my understanding certain tests are run on boot by the mobo to check everything is in order. Not much though otherwise it would take ages. TBH it is unlikely that with many missing pins it would actually boot. The mobo would simply not allow it because the cpu will tell it not to. This why the bios is very important. Without a bios the cpu cannot function. It simply doesn't no what to do.

If your wondering I study elecrical and electronic engineering and this is my very basic second year stuff. We also have a small FAB but its really old. We had to produce some functioning wafers in there. We are a little behind AMD and Intel on our node process though. We can manufacture at 65micro meters so our transistor sizes are more the 1000x larger than AMD or Intels. Still need a very powerfull microscope to see them though.
a b à CPUs
January 9, 2009 6:16:41 PM

gpippas said:
940 pins on a Phenom 2 I guaruntee you that 70% of those pins are used for testing. They are used in manufacture to make sure it works. They are also used by the bios of the mobo to check things are in order. One pin is cabanle of doing many different things.

You'll be surprised how many you can break and these things will still work. I would suggest the OP tries pulling them all out 1 at a time and let us no just how many it takes. Smiley face not necassary.


Given modern CPU clock rates, aren't many of these pins just grounds? I know that for mobo traces, GHz frequencies make them act like transmission lines with distributed impedance effects and which is why you don't see them having sharp angles(acts like an antenna).
January 9, 2009 6:45:31 PM

Yeah there are quite a few ground pins. And yes they can act like antennas. TBH you can make an antenna out of a lot of household items. Metal coat hangers make a surprisingly good antenna. Many years ago I used to use 1 as the antenna for the car when the original got ripped off. These days most cars just have 2 metal strips in the back window as an antenna. Damage those and costs a fortune.
January 10, 2009 1:06:43 AM

70% pins unnecessary? That's a severe exaggeration. A modern CPU is far more complex than a 4-bit ALU. The limiter for a 4-bit ALU is that neither the ceramic chip package nor the silicon die can be microscopic, even if the actual transistors take up a microscopic amount of real estate. So your piece of silicon there can actually hold 100s or 1000s of ALUs, but the chipmaker only printed one and left 99.9% of the area blank. There is no such minimum limitation on the silicon area or the ceramic container for a modern CPU such as the PhII.

Several hundred of the 940 pins connect directly to the two channels of DDR2 memory. Perhaps a hundred or more connect to power or ground. That's because the CPU is rated for 125W, and it would be silly to supply 100 amperes (if @ 1.25V) across such a thin piece of metal.

No doubt some of the pins are unnecessary (they leave some room for the future) or not quite essential until you push the 125W envelope and beyond. But such pins number nowhere near the amount you suggested.
January 10, 2009 9:37:55 AM

I will admit 70% was an exageration but not by much. You find I also did state that the ALU was a very basic in comparison to a modern cpu. Quite literally about 25 years behind. Also you can't just work out the amp draw of a cpu using just P = VI. Its not as simple as that. No cpu would survive 100A being pumped through it anyway. Also how many power supplies do you know of that supply 100A let alone all of that through a 4 pin 12V cpu rail.

TBH i'm quoting my university lecturer. I currently do not have the knowledge to know what all the pins do on a modern cpu. However he does teach microprocessor so he can't be talking that much BS.
a b à CPUs
January 10, 2009 9:53:16 AM

Well it sounds more right than what I can make from a white paper.
a b à CPUs
January 10, 2009 12:25:09 PM

He accidently the whole thing ...
a b à CPUs
January 10, 2009 5:00:12 PM

gpippas said:
Yeah there are quite a few ground pins. And yes they can act like antennas. TBH you can make an antenna out of a lot of household items. Metal coat hangers make a surprisingly good antenna. Many years ago I used to use 1 as the antenna for the car when the original got ripped off. These days most cars just have 2 metal strips in the back window as an antenna. Damage those and costs a fortune.


Nah - just drill a hole in your back window and stick a couple coathangers in it :) .

Hey, maybe the OP should try that with his 2 broken pins - whaddya think, eh??
January 10, 2009 5:36:21 PM

I think my friend accidently his processor too!
January 10, 2009 5:50:35 PM

Outlander_04 said:
A G clamp is always helpful to install cpu's that resist .

Put the cpu over the socket and then put the clamp around the mb and cpu . Tighten them up . Really tight . Its guaranteed to force the pins in properly .
This method is so versatile you can even get intel cpu's into AMD boards for extra performance .
EXTRA performance

i laughed so hard my side hurt
a b à CPUs
January 10, 2009 8:06:44 PM

haha
January 11, 2009 1:40:09 AM

gpippas said:
I will admit 70% was an exageration but not by much. You find I also did state that the ALU was a very basic in comparison to a modern cpu. Quite literally about 25 years behind. Also you can't just work out the amp draw of a cpu using just P = VI. Its not as simple as that. No cpu would survive 100A being pumped through it anyway. Also how many power supplies do you know of that supply 100A let alone all of that through a 4 pin 12V cpu rail.

TBH i'm quoting my university lecturer. I currently do not have the knowledge to know what all the pins do on a modern cpu. However he does teach microprocessor so he can't be talking that much BS.

For trying to express credibility for saying 70% is not a big exaggeration, you do make a glaring mistake.

On P = V * I, we are in agreement, where I is the amperes of current.

If the amperage (I) from the 12V CPU rail to the CPU remained constant, the drop from 12V to 1.1-1.35V (or whatever Vcore) would indicate a colossal waste of 80-90% of the total power (P). Do MOSFETs give off 4+ times more heat than CPUs?

The power circuitry including MOSFETs is over 80% efficient, meaning that just 15 amps transmitted through the 12V rail generate at least 120 amps at 1.2 Vcore to the processor. Almost every modern PSU can handle 15A @12V, and physical 8-pin connections are rated for >20A...

But a core i7 processor by default has overcurrent protection logic that throttles the CPU when current exceeds 100A or power exceeds 130W. Even when this logic is bypassed and ratings exceeded for overclocks, i7's seem to survive.

I don't have concise diagrams for some socket layouts, but pages 28 and 29 of this PDF comprise the entire pin layout diagram of socket 939, which is very similar to the modern AM2+ socket; the pin codes are explained from page 45 onward.

http://www.amd.com/us-en/assets/content_type/white_pape...

Notice the huge block at H6 to AD24? Over 300 pins are devoted to core power supply and matching ground. Ground is staggered in between power pins just as 40 ground wires shield 40 data/power wires from each other in 80-pin HDD IDE cables that go into 40-pin connectors. Are all the cables/pins carrying at capacity? Certainly not, but almost all are necessary there to preserve signal quality.

Of the 939 total pins, about 55 are marked NC. 94% of the pins have a publicized function. There are 157 VDD (core power), 58 VDDIO (IMC power), and 270 ground (over 200 shielding power and critical I/O pins). How many VDD and matching ground you need would depend on device amperage, internal circuit balance/isolation, and tolerances for voltage instability.

70% of 939 is 657. You can take out all the NC pins, but would you want to knock out the entire core/memory controller power supply, all ground and inductance shielding pins, and 150 other data carrying pins? That's what an exaggeration sounds like. It seems fewer than 25% of the pins can be taken out without major consequence. Taking out 25% would halve the socket's maximum TDP, or remove all shielding, or any proportional mix of the two, while presumptively removing all extraneous ground.
January 11, 2009 3:22:25 AM

Is seriousness allowed on this thread? Isn't it counterproductive?

Perhaps we could create a list of 101 things you can do with a dead CPU?

1. Open the package and spread peanut butter on the wafer, then tell everyone that you've invited the Toxic Avenger over for lunch.
      • 1 / 2
      • 2
      • Newest
!