How do they make CPU pins?

[grrl]

CPU pins are made to very close tolerances from what I've read, but how are they made - forging or stamping, and what kind of machining is done to finish them?

 

[TitanDiddly]

I imagine they come off of wire stock, then are stamped into shape and plated.

 

[grrl]

Originally posted by: TitanDiddly
I imagine they come off of wire stock, then are stamped into shape and plated.

That would make sense.

 

[MrDudeMan]

i work with wires and devices with way tighter tolerances than CPU pins, and honestly it is not nearly as hard as it may seem. i make preload wires for piezoelectric actuators on a daily basis and the tolerances are +/- 1 microinch, and i can do it by hand with very few tools. im sure the process to make and shape CPU pins is automated, which probably gives it somewhere in the +/-.1 microinch tolerances. one thing that should be noted is the length of the pins...they make them as short as possible because any wire or pin switching high current acts like a huge antenna and it radiates noise into everything around it.

 

[TitanDiddly]

Originally posted by: MrDudeMan
i work with wires and devices with way tighter tolerances than CPU pins, and honestly it is not nearly as hard as it may seem. i make preload wires for piezoelectric actuators on a daily basis and the tolerances are +/- 1 microinch, and i can do it by hand with very few tools. im sure the process to make and shape CPU pins is automated, which probably gives it somewhere in the +/-.1 microinch tolerances. one thing that should be noted is the length of the pins...they make them as short as possible because any wire or pin switching high current acts like a huge antenna and it radiates noise into everything around it.

How much current could a single CPU pin be handling?

 

[PsYcHoCoW]

according to p4 6xx datasheet, there are 226 Vcc pins and 273 Vss pins... Icc(max) is 119A for an extreme edition processor at 3.73GHz ; that means 525mA per Vcc pin, if the load is spread ideally... ( ftp://download.intel.com/design/Pentium4/datashts/30638203.pdf )

 

[Fox5]

Originally posted by: TitanDiddly

Originally posted by: MrDudeMan
i work with wires and devices with way tighter tolerances than CPU pins, and honestly it is not nearly as hard as it may seem. i make preload wires for piezoelectric actuators on a daily basis and the tolerances are +/- 1 microinch, and i can do it by hand with very few tools. im sure the process to make and shape CPU pins is automated, which probably gives it somewhere in the +/-.1 microinch tolerances. one thing that should be noted is the length of the pins...they make them as short as possible because any wire or pin switching high current acts like a huge antenna and it radiates noise into everything around it.

How much current could a single CPU pin be handling?

Maybe enough to interfere with other cpu pins?

 

[MrDudeMan]

Originally posted by: TitanDiddly

Originally posted by: MrDudeMan
i work with wires and devices with way tighter tolerances than CPU pins, and honestly it is not nearly as hard as it may seem. i make preload wires for piezoelectric actuators on a daily basis and the tolerances are +/- 1 microinch, and i can do it by hand with very few tools. im sure the process to make and shape CPU pins is automated, which probably gives it somewhere in the +/-.1 microinch tolerances. one thing that should be noted is the length of the pins...they make them as short as possible because any wire or pin switching high current acts like a huge antenna and it radiates noise into everything around it.

How much current could a single CPU pin be handling?

why the confused face? 22 gauge wire can handle 200,000 amps for a short period of time. when you turn your light on there are thousands of amps going through your lightbulb for a very short amount of time. i could go on, but the point is the size of the pin doesnt have much to do with its ability to handle enough current to cause noise. the length of a pin is more than enough to radiate enough energy to cause noise in other parts of the circuit.

 

[MrDudeMan]

Originally posted by: PsYcHoCoW
according to p4 6xx datasheet, there are 226 Vcc pins and 273 Vss pins... Icc(max) is 119A for an extreme edition processor at 3.73GHz ; that means 525mA per Vcc pin, if the load is spread ideally... ( <a target=_blank class=ftalternatingbarlinklarge href="ftp://download.intel.com/design/Pentium4/datashts/30638203.pdf">ftp://download.intel.com/design/Pentium4/datashts/30638203.pdf</a> )

good research. the traces on the motherboard can handle a heck of a lot of current in short bursts, and a moderate amount of sustained current. some of the PCBs i use at work look very weak (power wise) but are rated for 200A for < 200uS, and the traces are no wider than a few millimeters.

 

[sdifox]

on a somewhat related slant, are we gonna end up with 10 layer PCBs? It's getting very tight and integration has its limits.

 

[Spinne]

Nah, you can do it indefinately! ;-)