Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Surface Mounted Capacitors on PC CPU

Status
Not open for further replies.

victor6799

Junior Member level 2
Joined
Dec 6, 2016
Messages
20
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
108
My apologies I know that this does exactly apply to analog design but was hoping if any of you experts out there could shed some light on the following question about the Intel CPU.

I would like to know what is the purpose of the surface mounted capacitors found directly on the top of a CPU such as the Intel i3 2120 or the one depicted in the photo ? Does this type capacitor directly feed the internal circuitry of the CPU with a charge ? Perhaps directly to a specific layer on die of this Intel CPU ? If yes does a SMC and its charge ie voltage somehow follow a trace directly to a location a particular layer on die ?

IMG_1760.JPG
IMG_0653.JPG
 

This capacitors are called as On-die capacitors. You can google it for more data. Above certain frequency the on-board capacitors will be in effective due to unavoidable mounting inductance. To support decoupling at frequencies above the on-board capacitors effective frequency we need on-chip capacitors. These will be mounted very close to rail (intended to decouple).
 

This capacitors are called as On-die capacitors. You can google it for more data. Above certain frequency the on-board capacitors will be in effective due to unavoidable mounting inductance. To support decoupling at frequencies above the on-board capacitors effective frequency we need on-chip capacitors. These will be mounted very close to rail (intended to decouple).

Hello Mali. Thank you for the response. Can you tell me when you say "rail" do you mean the voltage rails ? If yes then could it be possible that a specific surface mounted capacitor or on die capacitors as you have described provides the decoupling features to a particular rail on a specific layer of the die ? I would imagine that each layer on die of any IC has a voltage rail yes ?
 

These capacitors provide a low impedance supply voltage to key areas of the chip. In general the inner layout is comprised of individual distinct cores or other complex circuits, but the wafer metalization process do not ensure sufficient charge capacity as an external component can do, which can grow up its capacitance with its height profile; a capacitor made on silicon is much more expensive if compared to any other technologies.

I have been repairing some industrial PCs form a manufacturer which do not assemble heatsinks with these 4 spacers on each corner as shown in the 2nd picture, but are more similar to the 1st picture. They took in consideration the gap between the CPU and Capacitors. I have been saving some mainboards from scrap with minor defects, but still usable, but it is a little annoying when I was trying to mount my own 'homemade' computer with them. I asked myself why didn't they simply covered them with some resin; perhaps due to termal issues.
 

I've done chip designs where things like very low jitter
requirements demanded close-in decoupling (else the
ESL of package leads and board would kill the high
frequency usefulness). We engineered the package
from CGA lands up to the lid, to get best power plane
inductances and so on. All 4 sides had multiple high-Q
capacitors attached on the topside.

Chips with high speed serial I/O (which is most modern
processors) need very low jitter and thus very low at-chip
power supply ripple in order to get link bit-error-rate
reliability.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top