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.

Why do we do dummy metal filling in layout?

Status
Not open for further replies.

s3034585

Full Member level 4
Joined
May 24, 2004
Messages
226
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,296
Activity points
2,087
hi
why do we do dummy metal filling in layoug (mixed signal/ analog/digital). can we use metal directly to do metal filling instead of dummy metal. How this metal has to be terminated, i know that we cant leave it floating but i am not sure about termination of it.
can you pls let me know about this.
thanks in advance
tama
 

dummy metal filling

I think it is used to flatten the surface of layout.
 

Re: dummy metal filling

consider a layout of two transistor...we should emphasize the importance of maintaining the same environment on the two side of the axis symmetry.

If one metal line passing over only one of the transistor indeed degrade the symmetry and incresing the missmatch....

Replica must be produceon the other transistor side,
 

dummy metal filling

it is a requirement from foundry.
 

Re: dummy metal filling

actually the dummuy can be one more transistor.
why we will use dummy is, suppose in anolog or digital sometimes the signal will mismatch with other device
as they are cmos devices somtimes there will not be balence in the circuit
so
for that purpose you need to use dummy so that the signal passage will be good.
a mosfet with all terminals grounded can be used as a dummy device.
so that there will not be any signal mismatching.
thank you.

Added after 2 minutes:

not only the dummy device but also we can use centroid method for the application
 

dummy metal filling

There is alot of usage in dummifications,
We can dummified the Metals, the VIAS(contacts between metals), and even the base layers such as diffusions, polys and nwells.... WHY? All above that mentioned is the answer, I will add on a few more, as during fabrication, materials are doped layers by layers, ....

Exclude the substrate/bulk... the other areas that didnt doped by materials that use for current flows is what we called the Field Oxide (fox) or ILD. This layer are soft, we need to dummified some metals enable to strengthen them, this is because when planarization phase in fabrication will cause the soft area tend to have some "valleys".
For better understanding, imagine dummy Metals as BRICKS, and the soft areas as SIMEN.
 

Re: dummy metal filling

Hi Guys
Thanks for all your replies. It was helpfull.
I would like to know one more thing that is after adding these dummy layers we cant keep them floating we need to connect them to some potential. What is that potential either vss or vdd. Do we need to add blockage layer also for all metals.
Thanks in Advance
Tama
 

dummy metal filling

IT depends , for my case I am designing chipsets here, floating metals and vias its ok for my case...

It wont flag those opens violations,....
 

Re: dummy metal filling

The metal dummy filling is required to meet the reliability by foundry. So ask the foundry for the rule.
 

dummy metal filling

there are many reasons for this.

one of them is like in analog circuits for the better performance of the devices in ckts like differential pair and current mirrors the surrounding environment should be the same or else the paracitic capacitance offered by that material on one side of the mosfet in a differential pair is different for the other mosfet in the same differential

one more reason is etal density issues.for a particular process technology it is pre defined that this metal should contribute this % of the total metal in the full chip

one more is some signals where box schelding is required we cover it by dummy metals and just tie them tot vdd or gnd
 

Re: dummy metal filling

Due to CMP(chemical mechanical processing) on metal, with dummy metal you'll have
small variation(metal loss). This is very important especailly for deep submicron technology since your metal thickness is quite small.
The common practice for dummy metal is to connect it to Vss or Vdd. Leaving it floating may cause crosstalk noise problem.
 

Re: dummy metal filling

s3034585 said:
hi
why do we do dummy metal filling in layoug (mixed signal/ analog/digital). can we use metal directly to do metal filling instead of dummy metal. How this metal has to be terminated, i know that we cant leave it floating but i am not sure about termination of it.
can you pls let me know about this.
thanks in advance
tama

Hope I can make it in a systematic way:

metal filling helps in two major issues

1) dishing in CMP process, which "bg21359" has been clearly explained :)

2) antenna effect occured during process step with plasma environment. This effect can cause gate oxide breakdown.

Simply... more metal there.... more antenna...... less energy abosrbed per metal.... less chance for the gate oxide being breakdown.

For detail, I suggest you do a search "antenna effect" in google, you will know it clearly.

Hope this help
Scottie
 
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top