Designer’s Guide Community

aaron_do

Senior Fellow

Offline

Posts: 1398

Via implementation
Jun 22^nd, 2011, 3:14am

Hi all,

I was wondering how a VIA on an IC actually looks like. Is it the same as a via on PCB where the VIA is actually tube-like rather than a solid cylinder?

If the VIA is tube like, I expect that if we were to take a bird's eye view of a metal with VIAs in it, we would see many holes in the metal. This could increase the resistance of the metal and potentially reduce its current handling capability. Does that sound right?

thanks,
Aaron

Back to top

there is no energy in matter other than that received from the environment - Nikola Tesla

IP Logged

RFICDUDE

Community Fellow

Offline

Posts: 323

Re: Via implementation
Reply #1 - Jun 22^nd, 2011, 3:46am

The modern copper vias tend to look like inverted cones (bigger on top than on the bottom. And they are solid (not hollow like unfilled pcb vias), but the material is not pure copper so the resistance may be a bit higher.

You can find some photo examples of IC cross sections showing vias and interconnect metal by performing image searches for

"company x" "cross section"

Back to top

IP Logged

aaron_do

Senior Fellow

Offline

Posts: 1398

Re: Via implementation
Reply #2 - Jun 22^nd, 2011, 6:28am

thanks for the info. I have one issue though, according to my process data, the VIAs of a particular metal are significantly more resistive (several times) than the same metal in the metal layer. That's why I came to the conclusion that they must be hollow. However, a cone-like structure can't seem to account for the large discrepancy in the resistance. Any ideas?

thanks,
Aaron

Back to top

there is no energy in matter other than that received from the environment - Nikola Tesla

IP Logged

rfcooltools.com Senior Member Offline Posts: 159	Re: Via implementation Reply #3 - Jun 22^nd, 2011, 11:50am Aaron_do, I found this graphic in the MIT fastcap software user guide on page 34. �Thought I would share. http://rfcooltools.com
Back to top	fastCapViaPick.PNG IP Logged

loose-electron Senior Fellow Offline Best Design Tool = Capable Designers Posts: 1638 San Diego California	Re: Via implementation Reply #4 - Jun 22^nd, 2011, 2:23pm have a look here: http://edfas.asminternational.org/portal/site/edfas/Photo2002/ aluminum and copper via interconnects between layers and then connections at the bottom layer of metal are all different. Got some more specifics and perhaps we can be of help
Back to top	Jerry Twomey www.effectiveelectrons.com Read My Electronic Design Column Here Contract IC-PCB-System Design - Analog, Mixed Signal, RF & Medical IP Logged

aaron_do

Senior Fellow

Offline

Posts: 1398

Re: Via implementation
Reply #5 - Jun 22^nd, 2011, 6:59pm

Hi all,

I have checked two different vias. One is going from Cu to Cu, and the other is going from Cu to Al (top layer). The one going from Cu to Cu has about 7 times higher resistivity than the Cu metal layer, while the one going from Cu to Al has about 18 times higher resistivity than the Al metal layer! Unless the cone really goes to a single point, I can't see how a conical shape can explain such huge increase in resistivity. Any ideas?

thanks,
Aaron

Back to top

there is no energy in matter other than that received from the environment - Nikola Tesla

IP Logged

aaron_do

Senior Fellow

Offline

Posts: 1398

Re: Via implementation
Reply #6 - Jun 23^rd, 2011, 12:48am

Hi all,

I have figured out the problem. It seems there were two issues,

1. The process data I had was wrong and the resistivity of the via is only 9 times (not 18 times) that of the metal).

2. More importantly, the resistance of a VIA seems to take into account the height of the layers it connects to. This was the main reason. EDIT: It probably includes half of the height of the VIA, and plus the fact that the VIA is effectively a corner.

cheers,
Aaron

Back to top

« Last Edit: Jun 23^rd, 2011, 1:59am by aaron_do »

there is no energy in matter other than that received from the environment - Nikola Tesla

IP Logged