Lapping kit and top of my Barton

G

Glow

Supreme [H]ardness
Joined
Jul 20, 2002
Messages
4,834
I know some people lap thier processors but how do I got about lapping it? What grains do you use? Heatsink isn't a problem but I dunno about the processor. I'm just paranoid also about scraping the black part around the metal.
 
Originally posted by Glow
I know some people lap thier processors but how do I got about lapping it? What grains do you use? Heatsink isn't a problem but I dunno about the processor. I'm just paranoid also about scraping the black part around the metal.
Click to expand...

lapping the bare core is nuts... people who lap processors are usually lapping P4's which have the metal caps protecting the core.
 
I've heard of a few ppl using goo gone (or something of the sorts) to get rid of that weird sticker-like thing on barton cores, then cleaning it with isopropyl alcohol. Temperature difference is slight, though, more noticeable if your heatsink is also lapped (still not very, though). Other than that, it'd be risky to lap an AMD core, though if you had the money, you could try and tell us how it goes.
 
well, you have a VERY GOOD chance of killing the barton, but on the other hand, you MAY get 1 degree difference..... definatly not worth the risk in my opinion
 
people lapping AMD cores usually did it when there was print on the core (such as the palomino cores). They were slightly bumpy. You dont really have to on new cores.
 
People who did it on the palomino and down processors were still nuts. That metal cap isn't remarkably thick, nor is it easy to precisely lap the whole surface of something that small.

Risk/reward ratio is totally out of whack..
 
Originally posted by kleptophobiac
People who did it on the palomino and down processors were still nuts. That metal cap isn't remarkably thick, nor is it easy to precisely lap the whole surface of something that small.

Risk/reward ratio is totally out of whack..
Click to expand...
i can agree to that my friend
 
Originally posted by kleptophobiac
People who did it on the palomino and down processors were still nuts. That metal cap isn't remarkably thick, nor is it easy to precisely lap the whole surface of something that small.

Risk/reward ratio is totally out of whack..
Click to expand...

What metal cap?
 
Do you really think that the shiny thing you're looking at in the center of the chip package is really the core? Nope.

The core is sandwiched between a layer that bonds to the conductive pads on the bottom of the silicon substrate (the plastic PCB) and a metal cap to protect the core. Yes, the metal cap is a ton smaller than Intel's IHS, but the P4 has a metal cap over the chip itself in addition to the IHS as well.

If there were no metal cap, the tiniest piece of dust (or thermal paste) would destroy the processor the moment it leaves the clean room.
 
Originally posted by kleptophobiac
Do you really think that the shiny thing you're looking at in the center of the chip package is really the core? Nope.

The core is sandwiched between a layer that bonds to the conductive pads on the bottom of the silicon substrate (the plastic PCB) and a metal cap to protect the core. Yes, the metal cap is a ton smaller than Intel's IHS, but the P4 has a metal cap over the chip itself in addition to the IHS as well.

If there were no metal cap, the tiniest piece of dust (or thermal paste) would destroy the processor the moment it leaves the clean room.
Click to expand...

So how does the core chip if it's got a metal cap on it? I was under the impression that it was just silicon that you saw. I know the microscopic electronic devices aren't exposed to the air, they're embedded in silicon. If they were exposed and had a metal cap over them, the metal would also destroy the electronics.
 
Interesting thread...

I have been kicking around the idea of lapping the processor on the GPU I see on NVIDIA and ATi boards. I suppose it could be easily done the same way you lap a HS. Using the same rationale for doing the HS it seems just as reasonable to do it to the GPU as well. Not sure I would be too keen on doing one of those tiny hunks of silicon on the AMD chips though...
 
I have lapped a few AMDs just for the h311 of it and for product testing purposes.
more specifically, I've lapped a tbird, a pally, two tbred 2100s and a 2500.
I used a product that is unfortunately not on the market as of yet (not my fault...the patent office is kind of slow this time of year).

anyway, I started with 1500 to see how flat the cores were and then dropped to 800 to flatten them up, then went to 1000, 1500 and 2000 for the final smoothing.

It helped a little but I cannot really say how much of the temp improvement was from the lapping as the aforementioned yet-to-be-released gadget produces a decent increase in cooling performance even when the core is not lapped.
 
Silicon is metal, sortof.

There are transisters etched into the silicon, with copper interconnects. The whole package is then sealed off with some adhesive and a metal plate. When you chip the "core", dust enters the silicon chip and kills it. Also, if you apply enough force to break the metal on top, who's to say it wasn't enough to break the silicon chip underneath?
 
Originally posted by kleptophobiac
Silicon is metal, sortof.

There are transisters etched into the silicon, with copper interconnects. The whole package is then sealed off with some adhesive and a metal plate. When you chip the "core", dust enters the silicon chip and kills it. Also, if you apply enough force to break the metal on top, who's to say it wasn't enough to break the silicon chip underneath?
Click to expand...
Just curious you on air or water cooling?
 
Originally posted by kleptophobiac
Silicon is metal, sortof.

There are transisters etched into the silicon, with copper interconnects. The whole package is then sealed off with some adhesive and a metal plate. When you chip the "core", dust enters the silicon chip and kills it. Also, if you apply enough force to break the metal on top, who's to say it wasn't enough to break the silicon chip underneath?
Click to expand...

Silicon isn't a metal. And if the top were metal, then it wouldn't chip. And also, I've used chipped processors before with no problems, and lots of other people have too.
 
Why do you think I listed silicon as a metal... sortof.

It is a semiconductor, the very property that makes processing possible. It resides in the murky world between metals and nonmetals. You can make transistors out of it in massive quantities, tens of millions in a hundred square millimeters.

Yes, metal can chip. Aluminum foil isn't the same as armor grade steel.

Also, using a chipped processor with no problems depends on the severity of the chip. If you don't bust off the whole protective layer over a section of chip, great. Not to mention, the chip doesn't actually take up the whole space of the core. The core is a protective package for the chip.

Let me go take a picture of what a naked (and non-functional) chip looks like.
 
Here, you disbelieving SOB. Here are naked chips that end up being used in inkjet printers. These end up being coated, at which point they can survive the non-cleanroom environment.

Sorry for the blurry picture, but you try taking pictures with a low-end digicam from three yeras ago in the dark of a shiny surface.

chips.jpg
 
here's a little perspective, those are smaller than your pinky fingernail.

Definitely not a UFO.

I've got a couple more wafers I can take pictures of. I even have an elusive 12" wafer.
 
Originally posted by kleptophobiac
Here, you disbelieving SOB. Here are naked chips that end up being used in inkjet printers. These end up being coated, at which point they can survive the non-cleanroom environment.

Sorry for the blurry picture, but you try taking pictures with a low-end digicam from three yeras ago in the dark of a shiny surface.

chips.jpg
Click to expand...

There's no need for your atittude. And I know what a semiconductor is, and I know how metals can behave. I'm a senior in college studying Materials Science & Engineering, so you don't have to act like I'm a moron. And as far as your aluminum vs steel argument, it's more likely for the steel to chip, seeing how aluminum never undergoes a ductile-to-brittle transition like steel does.

Now, is silicon the "metal" you were talking about that covers the top of the chip? You still haven't clarified what kind of metal it is or how you know it's there, which is all I was asking in the first place. If it's just a piece of metal on top of the transistors and interconnects, how is that not shorting all of them out?
 
Well, generally you don't bond the top of a chip exposed to the surface. You bond it to the substrate, and the blank side is exposed.

Upon further inspection of a Duron with a microscope, I don't see a cap at all. You are looking right at the back of the piece of silicon.

Do anything at all to that thing that violate the other side in the slightest, and you destroy a chip.

The situation is worse than expected.
 
Originally posted by dotZIP
I've heard of a few ppl using goo gone (or something of the sorts) to get rid of that weird sticker-like thing on barton cores, then cleaning it with isopropyl alcohol. Temperature difference is slight, though, more noticeable if your heatsink is also lapped (still not very, though). Other than that, it'd be risky to lap an AMD core, though if you had the money, you could try and tell us how it goes.
Click to expand...

Do not use goo gone on anything requiring thermal compound.

Goo gone has oil in that will get trapped in the minute pores of the surface and no amount of cleaning will get it out.
 
You must log in or register to reply here.
Back
Top