The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> isolation
https://designers-guide.org/forum/YaBB.pl?num=1178604466

Message started by aaron_do on May 7^th, 2007, 11:07pm

Title: isolation
Post by aaron_do on May 7^th, 2007, 11:07pm

Hi all,

sometimes Deep N-well is used to reduce substrate leakage. However, for the process i'm using, it must be added as an option and costs more. I was wondering how well standard N-well compares to deep N-well in terms of isolation, and whether standard N-well helps at all...

thanks,
Aaron

Title: Re: isolation
Post by sheldon on May 11^th, 2007, 8:45pm

Aaron,

The deep n-well that I am used to using is very different from the n-well
used as the body of a p-channel transistor. Please see page 15 of the following
slide set,

http://www.jazzsemi.com/docs/SiRF-06_Presentation.pdf.

A deep n-well is not used as the body of a p-channel transistor, it is used
to isolate the p-well used to create the body of an n-channel transistor
from the substrate. As a result, using a deep n-well makes the n-channel
much less sensitive to noise coupled through the substrate.

These structures are not interchangable, either you use a deep n-well and
the isolation is increased or you do not use a deep n-well and the body of the
n-channel is not isolated from the substrate.

Some additional comments, using a deep n-well improves the isolation of the
n-channel body and substrate. Depending on the frequency and the layout
there may still be isolation issues. That is,
1) There is still capacitive coupling so there may be leakage at high frequency
2) You need to make sure that the deep n-well does not contribute noise, that
is, the usual design practices to maintain a low-noise substrate: needing a
clean supply, plenty of substrate contacts[deep n-well contacts], guardbanding,
etc., apply to deep n-well layout
3) Whether or not you use a deep n-well in the design, you should still include
analysis of substrate coupling in the design flow

Best Regards,

Sheldon

Title: Re: isolation
Post by aaron_do on May 11^th, 2007, 10:29pm

Hi Sheldon,

thanks for the information. So you're talking about putting a p-well in a deep n-well. Actually I thought you could just use the deep n-well as a guard ring. Then any substrate noise would have to kind of go under the well. i.e it would isolate one area of the chip from the other. Not sure if they're used like that though.

Anyway thanks for the info. I guess i better try and characterize the substrate coupling and substrate leakage. Where would most of the substrate coupling come from?

thanks,
Aaron