Designer’s Guide Community

ywguo

Community Fellow

Offline

Posts: 943
Shanghai, PRC

Layout issue of MIM caps
May 27^th, 2004, 12:01am

Hi,

Traditionally, the top plate of cap is tied to critical point, like the input of opamp. That is because the bottom plate has parasitic capacitors, which will deteriorate the performance of that circuit.

Now, is it still a rule of thumb for the layout of MIM cap? I thought the bottom plate of MIM cap should have much smaller parasitic capacitance than that of PIP cap.

Thanks

Yawei Guo

Back to top

IP Logged

Paul

Community Fellow

Offline

Posts: 351
Switzerland

Re: Layout issue of MIM caps
Reply #1 - May 27^th, 2004, 12:32am

Hi Yawei,

the point is that you want the parasitic capacitance on the low-impedance side of your design. The bottom plate of MIM may have lower parasitics to substrate than PP, but it still has.

The question you have to ask is whether the top plate has more or less parasitics than the bottom plate. Connect the plate with less parasitics to the opamp input. If there is no higher-layer metal above your MIM caps, the bottom plate still has the dominant parasitics. In modern technologies, metal filling is often mandatory, so that you may end up with more parasitics at the top plate than at the bottom plate.

Paul

Back to top

IP Logged

ywguo Community Fellow Offline Posts: 943 Shanghai, PRC	Re: Layout issue of MIM caps Reply #2 - May 27^th, 2004, 7:43pm Paul, Thanks very much. Because I use MIM layer between the top layer and the second top layer, the bottom plate has dominant parasitic capacitance as what you said. By the way, is it still necessary to put nwell-shield under the MIM caps? ??? Yawei Guo
Back to top	IP Logged

Paul

Community Fellow

Offline

Posts: 351
Switzerland

Re: Layout issue of MIM caps
Reply #3 - Jun 25^th, 2004, 8:14am

Hi Yawei,

If you are using the top layers, take care about any metal lines running below. In modern technologies, the foundries tend to use metal filling. Floating metal below your MIM cap is definitely not what you want.

Regarding shielding, it depends on the noise in the layers below. Most efficient shielding is definitely achieved using the third layer from top, but you will get highest parasitic capacitance. It will solve the issue of metal filling on lower layers.

Paul

Back to top

IP Logged

ywguo

Community Fellow

Offline

Posts: 943
Shanghai, PRC

Re: Layout issue of MIM caps
Reply #4 - Jun 27^th, 2004, 9:30pm

Hi, Paul,

You are right. Any metal lines is forbidden under the MIM caps in the design rule. And I have never put dummy metal under the MIM caps, because I thought the behaviour of floating metal lines are not described clearly. I am not sure whether it couples with other signal lines or function as many antennas. ???

As to using the 3rd layer from top as shielding, I have never thought of that. If no floating metal lines is put under the MIM caps, is the 3rd layer more effective than N-well shielding if we don't consider the parasitic capacitance?

Thanks

Yawei Guo

Back to top

IP Logged

Paul Community Fellow Offline Posts: 351 Switzerland	Re: Layout issue of MIM caps Reply #5 - Jul 1^st, 2004, 12:32am Hi, I would say the closer the shielding the better it is (except for parasitic capacitance of course). But if your DR doesn't allow it, N-well shielding is the better way to go. If the foundry operates metal filling, I would recommend to put the blocking layers over the MIM area to avoid any bad surprises. Paul
Back to top	IP Logged