Designer’s Guide Community

Croaker Senior Member Offline Posts: 235	MOSFET Symmetry? Dec 12^th, 2006, 2:33am I read a line in Gray & Meyer that the MOSFET is not actually symmetrical in practice. Does anyone know why?
Back to top	IP Logged

Geoffrey_Coram

Senior Fellow

Offline

Posts: 1998
Massachusetts, USA

Re: MOSFET Symmetry?
Reply #1 - Dec 12^th, 2006, 6:02am

You might need to provide a little more context. For example, was G&M saying that actual devices won't be completely symmetric because of their environment? (well proximity effect, for example) Or even more basic, the source and drain areas might not be exactly equal in the real layout, because of fingers or diffusion sharing.

Of course, there are also devices that aren't meant to be symmetric, such as some LD- or HV-MOS devices, as well as those with "drain contact pull-back" for ESD robustness.

But G&M may have been referring to something more subtle, such as slightly different channel implants at the source and drain ends such that the saturation current is different when you flip the device around. I don't know if that happens in real life (or why), or if I'm just making it up.

Back to top

If at first you do succeed, STOP, raise your standards, and stop wasting your time.

IP Logged

Croaker

Senior Member

Offline

Posts: 235

Re: MOSFET Symmetry?
Reply #2 - Dec 12^th, 2006, 8:02am

They pretty much just said what I said, without offering any explanation.

I don't see how this is even possible, since there is no indication of the source or drain in a layout...so how could source and drain be manufactured differently? If you make a MOSFET and interchange the source and drain voltages, my understanding is that it should give the same results.

The most likely thing sounds like it could just be due to imperfections in the manufacturing process.

Back to top

IP Logged

vivkr

Community Fellow

Offline

Posts: 780

Re: MOSFET Symmetry?
Reply #3 - Dec 14^th, 2006, 8:11am

Hi Marc,

There are 2 different contexts where this difference can arise:

1. Most important is in high-voltage devices where the source and drain are designed differently. Typically, the
transistor has a gate which surrounds the source and is itself surrounded by the drain.

2. The ion implantation is done at an angle, which means that the 2 sides of a MOS are not quite symmetrical. This
is to prevent channeling. Now, the problem arises if you treat the right-hand side of transistor 1 as source, and the left
hand side of the other one as drain, and the 2 need to match. Razavi's book on CMOS analog design has some information about
this. I think it is the last chapter on fabrication.

Regards
Vivek

Back to top

IP Logged

Croaker Senior Member Offline Posts: 235	Re: MOSFET Symmetry? Reply #4 - Dec 14^th, 2006, 9:27am Regarding 2, who controls which side is the source or drain?
Back to top	IP Logged

didac

Senior Member

Offline

There's a million
ways to see the
things in life

Posts: 247
manresa,spain

Re: MOSFET Symmetry?
Reply #5 - Dec 14^th, 2006, 12:01pm

Hi.

I think you can't choose which terminal is the drain or source of the transistor. Instead you must use layout techniques to prevent or reduce the asymmetry like using pair of transistors gate-aligned or if you are forced to use parallel-gate transistors try to put dummy transistors in both sides of the pair to try ensure the two transistors see the same environment.(Sorry for my English, I'm from Spain and I'm not very used to write in English).

Back to top

IP Logged

vivkr

Community Fellow

Offline

Posts: 780

Re: MOSFET Symmetry?
Reply #6 - Dec 15^th, 2006, 12:40am

Croaker wrote on Dec 14^th, 2006, 9:27am:

Regarding 2, who controls which side is the source or drain?

Hi Marc,

As didac points out, there is no inherent definition of source/drain in case 2. Consider that different transistors on the very same die will experience this effect very differently, as there is no way of knowing a priori as to which side the ion implantation is done from. Moreover, depending on the orientation of a transistor, e.g. if it ends up seeing the ion beam parallel or perpendicular to its gate, the effect will also vary.

The only solution is to ensure that the direction of current flow is the same for any 2 transistors that must match. Note that this inherent asymmetry often presents a more severe problem than random mismatch. So, you have to lay out your devices very carefully to prevent this problem by choosing source/drain to lie in the same relative orientation for matching devices.

Regards
Vivek

Back to top

IP Logged

mc66

New Member

Offline

Posts: 6

Re: MOSFET Symmetry?
Reply #7 - Jan 13^th, 2007, 3:36am

you know when you layout ,you can dram a perfect symmetry MOS device

but when you manufcture it,it's a diffrent story,we can't align two layer exactly as layout,there misalign between
two diffrent layer,so in silicon,there is no really symmetry MOS

eg: In layout the space of cont to poly is same in both S/D,but in silicon,poly align to active,ct align to poly,and
misalign exist.so there are different space of cont to poly in S/D.

the design rule for guide layout take accout how much misalign of two layer and the cd variation of each layer.

Back to top

IP Logged

didac

Senior Member

Offline

There's a million
ways to see the
things in life

Posts: 247
manresa,spain

Re: MOSFET Symmetry?
Reply #8 - Mar 6^th, 2007, 2:51am

Recently as I was reviewing "The art of Analog layout" I found a reference to angle ion implantation(which causes asymmetry in MOS drain and source). It explains from a physics point of view so I need to review it carefully. The article its:"Ion Implantation in Semiconductors-Part I
Range Distribution Theory and Experiments",James F.Gibbons. It can be found at the IEEE ieexplore webpage.

Hope it helps

Back to top

IP Logged