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.

Which centroid matching method is good ?

Status
Not open for further replies.

sanjaytamne

Newbie level 6
Joined
Jul 21, 2007
Messages
11
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,363
centroid matching

I want to know which of the following centroid method is good and why?

A B B A
B A A B

or

A B A B
B A B A

OR

A A B B
B B A A
 

Re: centroid matching

the second one is not a common centroid but i choose it among the three.
but if you are to ask me between 1 and 3, option 1 is better because it has a better matching than the latter...
 

Re: centroid matching

The first one is preferred, because the axis of symmetry is in between the B's where ABBA is divided into 2 mirror image halves (AB - BA). However, dummies must be placed on each end of A's, coz the segments of this device occupy both ends of the array.
 

Re: centroid matching

Save your strength, for a small number of transistors using common centeroid is a waste of effort. it is more important to have symmetrical wiring since the effect of asymetrical nearby structure is more measureable then the differeces in doping level between the transistors.

this my private opinion, you can find others that thinkt the same. but there plenty of discussion - check what your boss believe :).

Nimrod Ben-Ari
 

centroid matching

I agree with Ben-Ari to a lot of extent, as in small number of devices, what matters most is the interconnect parasitics.

--cmos_dude
 

Re: centroid matching

if the device size is small, just take it easy:D

For the centroid layout, the second one is not,just select the first.
 

centroid matching

the second is better, i think. In opa layout, i design so
 

Re: centroid matching

The second one is best. It is more different in the same area, but the total differential effect coefficient is equal to the first one and the last one.
In a more clear way to say is that every four element array nearby of the second one are centroid symmetry.
 

Re: centroid matching

it's the first one. it follows the 4 rules of common-centroid layout.
 

centroid matching

I think the first one is best. As some friends said, the second one's wiring may do harm to the symmetry.
 

Re: centroid matching

wwwww12345 said:
I think the first one is best. As some friends said, the second one's wiring may do harm to the symmetry.


mirror symmetry allows mismatches due to process effects (implant angle shadowing, mask alignment errors, etc).
always use step symmetry (source-drain orientation is the same) to avoid these problems... wiring imbalance doesnt affect much the integrity of the the design as long as the devices are matched very well in the layout; but of course, it would be better if all are matched.. :)
 

Re: centroid matching

wiring imbalance doesnt affect much the integrity of the the design as long as the devices are matched very well in the layout; but of course, it would be better if all are matched..

i agree with this. there are designers who use the second one just to save the use of dummy devices but it's better to use the 1st one regardless of wiring imbalance. the 1st set agrees with coincidence, symmetry, dispersion and compactness. all of them may experience the same environment.
 

Re: centroid matching

For me, the first one is good. It covers the x and y axis matching.
 

centroid matching

yep,
ofcourse, the first one is better.
but it does n't make any diffenence in case of small size transistors.
 

centroid matching

yes i also think first one is best
 

centroid matching

can anybody explain how 2nd one is better than first
 

Re: centroid matching

The second one is better from my point of view
I agree with the people said that there is no need to do all of this stuff for tiny devices
It should be considered for diff pair & current mirrors espcially for short channel technologies
best regards,
Rania
 

centroid matching

I prefer to C because the matching is good enough and it helps to save area if drain/source is sharing.
 

Re: centroid matching

i think second one is better approach.
 

Re: centroid matching

"I prefer to C because the matching is good enough and it helps to save area if drain/source is sharing".

The option C doesnt provide good matching. Option A is the best in my opinion
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top