Link to PPT - mechatronics

runmidgeΤεχνίτη Νοημοσύνη και Ρομποτική

20 Οκτ 2013 (πριν από 4 χρόνια και 20 μέρες)

64 εμφανίσεις

Paper review:

Fractional Order Plasma Position Control of the
STOR
-
1M
Tokamak


Outlook of FOC in Plasma Etching:

Challenges and Opportunities

Zhuo

Li

PhD student, Dept. of ECE, USU.

zhuo.li@aggiemail.usu.edu


References


[1].
Shayok

Mukhopadhyay
,
YangQuan

Chen,
Ajay
Singh
and Farrell
Edwards
,
“Fractional
Order
Plasma Position Control of the STOR
-
1M
Tokamak
”, 48th IEEE
CDC, Dec, 2009, pp.422
-
427.


[2].
Mukhopadhyay
,

Shayok
, "Fractional Order Modeling and Control: Development of Analog
Strategies for Plasma Position Control
of the
Stor
-
1M
Tokamak
" (2009).
All Graduate Theses and
Dissertations.
Paper 460
. [online available], http
://
digitalcommons.usu.edu/etd/460


[3].
M.
Emaami
-
Khonsari
, “
Modelling

and Control of Plasma Position
in the
STOR
-
M
Tokamak
,”
Ph.D., University of Saskatchewan, Saskatoon
, April
1990
.


[4]. Shane Lynn,
“Virtual
Metrology for Plasma Etch
Processes”,
PhD thesis, Electronic Engineering
Department,
National Univ. of Ireland
.


[5] John V. Ringwood
,

Shane
Lynn,
Giorgio
Bacelli
,
Beibei

Ma
,
Emanuele

Ragnoli
, and Sean
McLoone
, “Estimation
and Control in Semiconductor
Etch: Practice
and
Possibilities”, IEEE TRANS
ON SEMICONDUCTOR MANUFACTURING, VOL. 23, NO. 1, FEBRUARY
2010


[6]. Lynn
Fuller,

Plasma Etching”
, [lecture slides],
Microelectronic Engineering, Rochester Institute
of Technology
.


[7].
Henri Janseny, Han Gardeniers, Meint de Boer, Miko Elwenspoek and Jan Fluitman,

A
survey
on the reactive ion etching of silicon in
micro
-
technology”
,
J. Micromech. Microeng. 6 (1996), 14

28
.


[8]. Lab modules, webpage, [online], http://matec.org/ps/library3/secure/modules/047/,
[Mar.16.2012]




Slide
2

The Physical
S
ystem


Tokamak
: is a device using a magnetic field to confine a
plasma in the shape of a
“doughnut”. [Wikipedia.org]

Slide
3

Fig3
-
1. The schematic of
Tokamak

as a
transformer.[1]

Fig3
-
2. The STOR
-
1M
Tokamak

in USU. [1]

Bank Current Waveforms


B
T

-

Toroidal

field
bank


I
Oh

-

Ohmic

heating
bank


I
V
e

-

Vertical equilibrium bank


I
H
c

-

Horizontal
compensation bank


I
V
c

-

Vertical
compensation bank

Slide
4

Fig4. The bank current waveforms of STOR
-
1M. [1]


Plasma position estimation mechanism [3]


Four magnetic “pickup” coils measure the

magnetic field produced by the
toroidal


plasma current.


By comparing

the strengths of this measured magnetic

field one can estimate the position

of the current.


Proposed technique in the paper


Ratio of the voltages


𝐵
𝑜
=
𝛼
𝑜

𝑜
=
𝜇

𝐼

2
𝜋


𝐵
𝑖
=
𝛼
𝑖

𝑖
=
𝜇
𝑖
𝐼

2
𝜋

𝑖














(
𝛼
𝑜

𝛼
𝑖
)



𝑖
+

𝑜
=
100


E.g.
𝑉

𝑉
𝑖
=
𝐵

𝐵
𝑖
=



𝑖
=
5
.
6667





(
Must

hit

the

wall
)



Fig5
-
1. The Plasma
p
osition estimation system.[3]

Measurement Mechanism

Slide
5

Fig5
-
2. Proposed position estimation approach. [3]


The transfer function



𝑌

=

2
.
84

3
+
1
.
14𝑒7

2
+
1
.
12𝑒14
+
4
.
5𝑒20

4
+
1
.
51𝑒4

3
+
3
.
9𝑒13

2
+
4
.
3𝑒17
+
2
.
15𝑒21

(

)



First order plus delay model approximation



𝐺

=

𝑇
+
1
𝑒


=
0
.
2096
0
.
0864

+
1
𝑒

0
.
0007




Plasma
P
osition Modeling

Slide
6

Controller

Kp

Ki

Kd

order

FO
-
PI

170.2649

32.0719

0

0.7425

ZN
-
PID

706.584

714.285

0.00035

1

Fractional Order
C
ontroller

Slide
7

Table: CONTROLLER
PARAMETERS FOR THE STOR
-
1M TOKAMAK

Fig8
-
1. Position
control
results. [3]


Controller parameters





Results
and
comparison (on emulator)


Fig8
-
2. Position
control
results. [3]


FO
-
PI controller is better than the
ZN
-
PID controller
in
terms of response time, control effort and
steady state
error.

Conclusion

Slide
8


Plasma etching process in semiconductor
manufacturing


Etching variables hard to measure


Real
-
time control hard to achieve


Measurement technology in the literature [5]


Virtual metrology [4]


Optical emission spectroscopy (OES)


Mass spectrometry


Plasma impedance monitoring


Etc.

Outlook
-
challenges

Slide
9

Fig9. OES. [4]

Plasma
E
tching
-

Intro


Etching outcome and profile


Isotropic

(non
-
directional removal of material from a substrate)


Anisotropic

(directional)


Slide
10

Ideal etch

Fig10
-
1.
No process is ideal, some
anisotropic plasma etches are close.
[6]

Poor etch

Fig10
-
2.
One
-
run multi
-
step RIE process. Top left: after
anisotropic etching the top Si of an SOI wafer. Top right: after
etching the insulator and sidewall passivation. Middle left:
during isotropic etching of the base Si. Middle right: after
isotropic etching the base Si. Bottom: typical finished MEMS
products.
[7]

The Plasma Etching Chamber

Slide
11

Fig11
-
2.
RIE

Process
Chamber
.
[8]

Fig11
-
1:
Typical parallel
-
plate
RIE

system.
[*]

Fig11
-
3:
Typical
RF sputtering
system.
[*]

Fig11
-
4.
Physical etch process
chamber
.
[8]

*
MEMSnet
®, https://www.memsnet.org/mems/beginner/etch.html

Controls in the Literature

Slide
12

Fig12.
Etch tool control possibilities with information
flow. [4]

Controls in the Literature

Slide
13


Run
-
to
-
Run (R2R) Control [a],[b],[c].


Predictive functional
control [4].


Neural network control


Etc.


[a], M
. Hankinson, T. Vincent, K.B.
Irani
, and P.P.
Khargonekar
. Integrated
real
-
time and
run
-
to
-
run
control of etch depth in reactive ion etching. IEEE T.
Semiconduct
. M
., 10(1):
121
-
130
, Feb. 1997
.

[b]. X.A
. Wang and R.L.
Mahajan
.
Articial

neural network model
-
based
run
-
to
-
run process
controller.
IEEE Trans. Components, Packaging, and
Manufacturing Technology
, Part C., 19(1):
19
-
26
, Jan. 1996
.

[c].
J.P. Card, M.
Naimo
, and W.
Ziminsky
. Run
-
to
-
run process control of a
plasma etch
process with
neural network
modelling
. Qual.
Reliab
. Eng. Int., 14(4):
247
-
260
, 1998.


Data


Outlook
-
Opportunities

Slide
14

Fig10. Endpoint mono
-
chromtor

output over four
preventative
maintenance (PM
)
cycles. [4]


Other
efficient “learning machines”


RVM


Other fitting methods


TLS fitting for “data boxes” (not point)


Interval computation tools (
IntLab
)


Dynamic VM


R2R VM


Fractional Order ANN based VM


Neuronal dynamics is inherently “fractional order



Fractional order
iterative learning control


Cognitive process control

Slide
15

Slide from
Dr.Chen’s

Lam Research Talk

Outlook
-
Opportunities


Dynamic Virtual Metrology in Semiconductor
Manufacturing


http://bcam.berkeley.edu/research/new_researchframes.html

Outlook
-
Opportunities

Slide
16

Slide from
Dr.Chen’s

Lam Research Talk

Thank you!


Q&A

Slide
17