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.

substrate connection in mixed signal circuit

Status
Not open for further replies.

John Xu

Member level 3
Joined
Jul 22, 2005
Messages
59
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,938
Hi,
i have a question on the substrate connection for mixed signal circuit with p-sub lightly-dopled submicron cmos process. In my mixed block, we have seperate analog vdd/gnd and digital vdd/gnd for crosstalk reduction from digital part.

I want to know the substrate connection, in my cicruit, i connected it to the analog gnd for which we believe it is most approviate for our cicruit. But i wonder if we can assign the substrate with different gnd. i.e., for analog part, we connect its substrate to analog gnd, for digital part, we connect it to digital gnd. To reudce the crosstalk between them, I put guardings to surround the sensitive analog and digiatl part, respectively, and laso, I put some heavy substrate ties strip between the analog and digital part. Seperate the analog sub and digital sub can avoid the body effect of the analog/digital part both. Is any risk in it?

Pls. comment!!!

Thanks in advance
 

Yes, you can do it. Always the substrate is tied to its own circuit block. And sometimes substrate can be tied to another clean ground line if there are
enough pins. But take care of other issue such as connection metal width and so on.
 

qtarch said:
Yes, you can do it. Always the substrate is tied to its own circuit block. And sometimes substrate can be tied to another clean ground line if there are
enough pins. But take care of other issue such as connection metal width and so on.

Hi, qtarch,
Thanks for the reply. You said " Always the substrate is tied to its own circuit block. " But in Razavi's "design of analog cmos integrated circuit", Page 664, it asserts that "of courrse, connecting the substrate to both GNDA and GNDD give rise to a low-resistance path between the two, defeating the purpoase of separating the analog and digital grounds." It seemed that he think the substrate is always tied to one gnd.How can I understand it?

Also you said"And sometimes substrate can be tied to another clean ground line if there are enough pins."Assume I have another pin can be assigned to substrate, we called this sub gnd. Do you really recommend me to connect the sub gnd and chip gnd respectively? Do you have any concern of "substrate bounce" effect or body effect?

Thanks
 

This depends greatly on the process. Do you have a well for the NMOS devices or is it simply the intrinsic substrate?

If the prior, it is easy to separate analog and digital ground and tie them separately to a GND pad. If the latter, it is significantly more complicated since the bulk of all NMOS devices is tied together (through the p-substrate).
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top