Transcript Under Frequency Relay Setting for Tie Line and Load Shedding of

Power control strategies evaluation of a
series resonant inverter for atmosphere
plasma applications
M. T. Tsai
Department of Electrical Engineering,
Southern Taiwan University, Tainan,
Taiwan
Outline
I. Atmosphere plasma Introduction
II. The research system configuration
III. Control system design
IV. Experimental results
V. Conclusion
I. Induction
Examples of industrial plasma applications：
• Low-pressure (vacuum pressure (< 10 mTorr), moderate pressure (~
1 Torr))
 Glow discharge plasmas: non-thermal plasmas generated by the
application of DC or low frequency RF (<100 kHz) electric field to the
gap between two metal electrodes. Probably the most common plasma;
this is the type of plasma generated within fluorescent light tubes.

Capacitively coupled plasma (CCP): similar to glow discharge plasmas,
but generated with high frequency RF electric fields, typically 13.56
MHz. These differ from glow discharges in that the sheaths are much less
intense. These are widely used in the microfabrication and integrated
circuit manufacturing industries for plasma etching and plasma enhanced
chemical vapor deposition.
Low-pressure
 Inductively coupled plasma (ICP): similar to a CCP and with similar
applications but the electrode consists of a coil wrapped around the
discharge volume which inductively excites the plasma.
 Wave heated plasma: similar to CCP and ICP in that it is typically RF (or
microwave), but is heated by both electrostatic and electromagnetic
means. Examples are helicon discharge, electron cyclotron resonance
(ECR), and ion cyclotron resonance (ICR). These typically require a
coaxial magnetic field for wave propagation.
Atmospheric pressure (760 Torr)
 Arc discharge: this is a high power thermal discharge of very high
temperature ~10,000 K. It can be generated using various power supplies.
It is commonly used in metallurgical processes. For example it is used to
melt rocks containing Al2O3 to produce aluminium.
 Corona discharge: this is a non-thermal discharge generated by the
application of high voltage to sharp electrode tips. It is commonly used
in ozone generators and particle precipitators.
Atmospheric pressure (760 Torr)
 Dielectric barrier discharge (DBD): this is a non-thermal discharge
generated by the application of high voltages across small gaps wherein a
non-conducting coating prevents the transition of the plasma discharge
into an arc. It is often mislabeled 'Corona' discharge in industry and has
similar application to corona discharges. It is also widely used in the web
treatment of fabrics. The application of the discharge to synthetic fabrics
and plastics functionalizes the surface and allows for paints, glues and
similar materials to adhere. A RF generator (20-100 kHz, 0-2.4 kV peak) is
used for discharges in DBD tube
 Capacitive discharge: this is a nonthermal plasma generated by the
application of RF power (e.g., 13.56 MHz) to one powered electrode,
with a grounded electrode held at a small separation distance on the order
of 1 cm. Such discharges are commonly stabilized using a noble gas such
as Helium or Argon.
II. The research system configuration
L
ig
Iin
D
iL
DQA
AC
+
Vin
_
Vg
Q
CO
RL
+
Q
DQC
QC
CQA
A
CQC
1
Vdc
RS
iR
NP : NS
+
VAB
Vdc
LR
-
-
QB
+
Lm
VP
-
DQB
CQB
QD
C
g
+
Vs
-
C
DQD
d
CQD
2
IR2110
Driver
1
K
IR2110
Driver
PDM/PWM
Select logic
UC3895
PWM controller

Fig. 1(a) PFC stage
VZ
PIC16F877
controller
Fig. 1(b) Inverter stage
The research system configuration
Leq
½u
Cg
Cg
Vcg
VZ
Cd
VZ
Vs
Cd
Cd
(b)
(a)
Fig. 2 Reactor model in different
operation status, (a) before plasma,
(b) plasma operation.
A simplified series resonant inverter
model
The research system configuration
• For the inverter output voltage and frequency at which the value of
voltage amplitude on capacitor C g can not reach the gas discharge value, VZ
the system has the maximum resonant frequency of f r max .
Cg Cd
1
f r max 
, Ceq 
(1)
Cg  Cd
2 LeqCeq
When the inverter output voltage and frequency at which the value of
voltage amplitude on capacitor C g reaches the gas discharge value, VZ ,
the system resonant frequency will be reduced to minimum situation
of f r min .
1
f r min 
,
(2)
2 Leq Cd
The research system configuration
• The value of VS can be shown as follows:
VS 1 
4

NS

cos ,
NP
2
Vdc
(3)
Thus, the voltage at the capacitor C g can be obtained as follows:
Vcg
VS

CZ 
CZ
1
1
C

,
d
2
2
 Leq Cd Cg  (Cd Cg )
( Leq CZ  1) C g
Cd C g
Cd  C g
,
(4)
The research system configuration
• Assuming VZ as the gas discharge starting voltage, then the switching
boundary frequencies at which gas discharge starts can be obtained as
follows:
f S max 
1
2
1 Cd VS
1
(  )
LeqCd CZ VZ
2
1 Cd 4 Vdc N S

( 
cos ), (5)
LeqCd CZ  VZ N P
2
The switching frequency of the inverter should meet the above equations,
and f S max is dependent on the parameters including transformer, plasma
reactor, and the values of the inverter input voltage and the inverter
output pulse-width.
III. Control system design
• A. PFC stage
• The PFC stage uses the UC3854 based average-mode controller to accomplish
fixed frequency current control with stability and low distortion. Unlike peak
current-mode, average current control accurately maintains sinusoidal line
current without slope compensation and with minimal response to noise
transients.
V AC
VO
Current sensor
UC3854N
1
16
2
15
3
14
4
13
5
12
6
11
7
10
8
9
Control system design
• B. Inverter stage
• The inverter has four stages, determined by the power
switching elements of the two legs. The stages in which two
diagonally opposite power switches are conducting are called
active [8]. On the contrary, the stages in which two switches
on the same site of power switches are conducting are called
passive. The switching of the leg can moves the inverter from
active stage to passive stage is called the leading leg (QC，
QD). The other leg which switches only from passive stage to
active is called the trailing leg (QA，QB).
Control system design
DQA
CQA Q
C
QA
A
NP : NS
-
+
DQD
CQB QD
VP
iR
VZ
QA
Cd
CQD
Vab
Cg
-
DQB
QB
iR
LR
VAB
B
Vab
CQC
+
VIN
iR
DQC
QB
QC
QD
t0
t1 t 2
t3 t 4
t5
t
Control system design
Vab
iR
CH 1
CH 2
A experimental result
Control system design
• For phase shift control and with the lossless snubbing capacitor,
the ZVS can be achieved in the leading leg for all the load
conditions, but can be achieved in the trailing leg only in the case
that the inverter operates with a lagging load current. For a RLC
series, it means the inverter switching frequency should be higher
than the load resonant frequency.
DQA
DQC
CQA Q
C
QA
CQC
+
VIN
A
LR
V AB
B
NP : NS
+
Cg
VP
-
VZ
-
DQD
DQB
QB
iR
CQB QD
Cd
CQD
t0  t1
Control system design
DQA
DQC
CQA Q
C
QA
CQC
+
VIN
iR
A
LR
V AB
B
-
CQC
QC
+
A
iR
LR
V AB
NP : NS
+
Cg
VP
-
VZ
-
DQD
DQB
QB
Cd
DQC
CQA
VIN
t1  t2
CQD
DQA
QA
VZ
-
DQD
CQB QD
QB
Cg
VP
DQB
B
NP : NS
+
CQB QD
Cd
CQD
t 2  t3
Control system design
DQA
DQC
CQA Q
C
QA
CQC
+
VIN
A
LR
V AB
B
-
-
CQC
+
A
iR
LR
V AB
NP : NS
+
Cg
VP
-
VZ
-
DQD
DQB
QB
t3  t 4
DQC
CQA Q
C
VIN
Cd
CQD
DQA
QA
Cg
VZ
DQD
CQB QD
QB
NP : NS
+
VP
DQB
B
iR
CQB QD
Cd
CQD
t 4  t5
Power control strategies
 PAM control
• The PAM controls the inverter input voltage by controlling the PFC stage
output voltage to achieve the adjusting inverter output power.
• Considering the power elements stress and wide voltage range, it is
preferred buck-boost scheme to boost scheme as the PFC stage.
• As the instantaneous electrode voltage should be large than the gas
breakdown level so as to form a sustain discharge procedure, thus the
inverter input voltage can not be lower than a certain level, so it leads to
a result that discharge power is difficult to less that half of the full range.
• A low voltage operation may be accompanied a partial or local discharge
due to inequality of the gas between the electrode and dielectric.
• The inverter output power factor decreases as the output increases, if the
operation frequency is constant.
Power control strategies
 PWM control
• The PWM controls the pulse width of the inverter output voltage to achieve the
adjusting output power.
• By shifting the phase difference of the control phase with respect to the standard
phase, the output power can be varied from full power to low power, therefore it
is feasible to regulate the inverter output power.
• Lossless snubbing capacitors can enable the inverter to perform ZVS function
when it operates with a lagging load current.
• For discontinuous load current or leading load current, the ZVS function can not
be achieved in the trailing leg, resulting in an increasing switching loss.
• Similarly to the PAM scheme, the inverter output power factor also decreases as
the pulse width increases in the PWM control when the switching frequency is
constant.
• A small pulse width tends to a discontinuous load current or leading load current,
which is adverse to the switching loss, thus it is disapproved for a low pulse
width control.
Power control strategies
 PFM control
• The PFM controls the frequency of the inverter output voltage to achieve
the adjusting output power.
• To realize zero voltage switching, the inverter output frequency should
be large than the load resonant frequency. Thus, one can see that the
inverter power fact should decline in low power range.
• The inverter output power will have a steep increase when the frequency
is approaching the load resonant frequency. This will increase the
difficulty to control the inverter output power stably.
• As the same as the PAM and PWM situations, it is difficult for the PFM
applied to the voltage-source series-resonant inverter to adjust the
discharge power to less than half of the full power[3], as the electrodes
voltage would be lower than the gas discharge breakdown voltage.
Power control strategies
 PDM control
• The PDM controls the output power by controlling the number of
inverter output voltage pulses, in other words, it is to repeat “run and
stop”, in accordance with the desired output power. For example, if a
working cycle represents a time interval of 40 working pulses, thus by
varying the number of these working pulses from 10 to 40 will have a
regulated output power from 25% to 100%.
• [1] has shown this method can work well over a range of pulse densities
from 3/30 to 1.
• The PDM scheme can have lower switching loss than other schemes as it
achieves quasi-ZCS and ZVS functions [3]. However, if a shorten rising
time when the next discharge period starts is desired, it was found useful
to apply pulses with reduced width to the plasma reactor during zeropower periods, which is used to prevent deionization of gas [4]. Thus, the
control signal will consist of full-width pulses during the discharge
period and reduced-width pulses during the zero-power period.
Power control strategies
UC3895 structure
Power control strategies
•
PDM control
PIC OUT
3895 PWM
PDM
Period
Ton
Period
Toff
T
PIC OUT
3895 PWM
PDM
Period
Ton
Period
Toff
T
PIC OUT
3895 PWM
PDM
PDM control signals
Power control strategies
•
PDM control
PIC Out Signal
PIC 輸出信號
QA
OUTA 輸出信號
UCC 3895
QB
QC
Q A驅動信號
QD
PDM implementation –method 1
Q D驅動信號
Power control strategies
•
PDM control
PIC Out Signal
Q
D
QA
PIC 輸出信號
CK
UCC 3895
A
B
C
D
A組 Q
D
Q
CK
OUTA 輸出信號
QB
QC
B組 Q
Q A驅動信號
QD
PDM implementation –method 2
Q D驅動信號
IV. Experimental results
•
•
CH1-20A CH2-200V M-5ms/div
Fig. 3(b) The source voltage and current for a load case of 1000W
IV. Experimental results
Vgs
CH1
Vds
CH2
•
•
Vgs
M：10us
CH1：5V CH2：200V
Leading leg switching performance
IV. Experimental results
Vgs
CH1
Vds
CH2
•
•
CH1：5V CH2：200V
M：10us
Lagging leg switching performance
IV. Experimental results
Vds
CH1
Ids
CH2
•
•
CH1：400V
CH2：5A
ZVS switching performance
M：5us
IV. Experimental results
V AB
iR
(400V/div)
(10A/div)
Vs
(5kV/div)
•
(10uS/div)
Fig. 4(a) The experimental waveforms of 50% PWM control
IV. Experimental results
VAB
iR
(400V/div)
(10A/div)
Vs
(5kV/div)
•
(10uS/div)
Fig. 4(b) The experimental waveforms of 75% PWM control
IV. Experimental results
(400V/div)
VAB
(10A/div)
iR
Vs
(5kV/div)
•
(10us/div)
Fig. 5(a) PFM control for 50 kHz switching frequency.
IV. Experimental results
VAB
iR
(400V/div)
(10A/div)
Vs
(5kV/div)
•
(10us/div)
Fig. 5(b) PFM control for 40 kHz switching frequency.
IV. Experimental results
VAB
•
( 400V/ div)
iR
(10A/ div)
Vs
(5kV/ div)
( 100us/ div)
Fig. 6(a) PDM control for a case of during operation at a pulse density of
25% (10/40).
IV. Experimental results
VAB
•
( 400V/ div)
iR
(10 A/ div)
Vs
(5 kV/ div)
( 100 us/ div)
Fig. 6(b) PDM control for a case of during operation at a pulse density of
50% (20/40).
V. Conclusions
As the experimental results, some conclusions can be made as follows:
• The only PAM control is not encouraged as it needs a complicated
front stage to achieve voltage regulation function, and is hardly used
to less that half of the full range due to the required gas breakdown
voltage level
• The PWM control can fulfill the full load range conditions. However,
a small pulse width tends to a discontinuous load current or leading
load current, which is adverse to the switching loss, thus it is
disapproved for a low pulse width control.
• The only PFM control is also not encouraged as it should be large
than the load resonant frequency to realize zero voltage switching.
Thus, one can see that the inverter power fact should decline in low
power range, and it is difficult to adjust the discharge power to less
than half of the full power as the electrodes voltage would be lower
than the gas discharge breakdown voltage.
• The PDM can work well over a range of pulse densities from 3/30 to
1, however, the environment temperature fluctuations should disturb
the stability of the inverter output power. To compensate this
influence, a hybrid control such as PDM plus PFM or PDM plus
PWM is suggested.
Thanks for your attentions