Proceedings of the IEEE 2002 Custom Integrated Circuits Conference (Cat. No.02CH37285)
DOI: 10.1109/cicc.2002.1012872
|View full text |Cite
|
Sign up to set email alerts
|

Understanding MOSFET mismatch for analog design

Abstract: This paper addresses misconceptions about MOS-FET mismatch for analog desi . V, mismatch does not follow a simplistic 1/( s " area) law, especially for widelshort and narrowllong devices, which are common geometries in analog circuits. Further, V, and gain factor are not appropriate parameters for modeling mismatch. A physically based mismatch model can be used to obtain dramatic improvements in prediction of mismatch. This model is applied to MOS-FET current mirrors to show some non-obvious effects over bias,… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

2
77
2

Publication Types

Select...
5
3

Relationship

0
8

Authors

Journals

citations
Cited by 56 publications
(81 citation statements)
references
References 19 publications
2
77
2
Order By: Relevance
“…Typical mean conditions were used, introducing independent Gaussian deviations of W , L, µ 0 and t ox with σ = 10%. These deviations generate in turn deviations in the crucial parameters V th , γ and β [7] which directly affect the previously described equations defining the processing. The results are summarized in Table II.…”
Section: Simulation Resultsmentioning
confidence: 99%
“…Typical mean conditions were used, introducing independent Gaussian deviations of W , L, µ 0 and t ox with σ = 10%. These deviations generate in turn deviations in the crucial parameters V th , γ and β [7] which directly affect the previously described equations defining the processing. The results are summarized in Table II.…”
Section: Simulation Resultsmentioning
confidence: 99%
“…Global variation does not related to circuit layout and device sizes, while local variation does [6,8].…”
Section: Introductionmentioning
confidence: 99%
“…A common limitation of both alternatives is that they consider only parasitic without tackling other layout aspects, e.g., process parameter variations, mismatching, thermal effects and substrate coupling. Especially for deep submicron (DSM) design, process variations are arguably as critical as parasitic, and ought to be considered early in the design flow [6].…”
Section: Introductionmentioning
confidence: 99%
“…An equation equivalent to (6) can be derived for ∆β i variations in series transistors. Variations in β and V T , in parallel transistors are calculated as in (3). The normalized mismatch current standard deviation is obtained: …”
Section: Current Mirror Topology and Mismatchmentioning
confidence: 99%
“…For the sake of simplicity, only variations in V T and β -as in (1)-will be considered in this paper. Nevertheless the analysis can be extended to other models to study the mismatch in CMOS transistors [1][2][3][4], or to obtain more accurate expressions for 2 T V σ , 2 β σ , considering, for example, the mismatch dependence on the distance between transistors [1,2]. In analog design, it is common to express mismatch in terms of ∆V T , ∆β, the difference between V T , β, of two matched transistors.…”
Section: Introductionmentioning
confidence: 99%