MOS-FET Based Marx Generator for Application to Electron

Download Report

Transcript MOS-FET Based Marx Generator for Application to Electron

Developments of fast response kicker
magnets for Rare-RI Ring
Akira Tokuchi
Pulsed Power Japan Laboratory Ltd.
Nagaoka University of Technology
Yoshitaka Yamaguchi
RIKEN Nishina Center
Akira Ozawa
Institute of Physics, University of tsukuba
 New Pulsed Power Company in Japan
 Established in May, 2009
 Our president has developed many
pulsed power machines for more than
30 years
President Akira Tokuchi
Conceptual design for Rare-RI Ring
Requirements and Technical problems
for Kicker Magnet PS
 Output Current
 Pulse Width
 Delay from Trigger
 Charging Time
Requirement
2kAp
200ns
←
<290ns
←
<200μs
←
Achieved
2kAp
1125ns
2.67μs
>10ms
Technical Problem 1: shortening of delay time from
trigger
Technical Problem 2: shortening of charging time
Technical Problem 1: shortening of delay time
from trigger
Conventional Block Diagram
550ns
<290ns
required
2.67μs
Faster thyratron gate unit was required
Technical Problem 1: Shortening of delay time form trigger
New Thyratron gate board
1.Faster Optical Link
(DC9351:hitachi)
2.Faster FET Drivers
(DEIC420:IXYS)
3.Faster MOS-FETs
(DE475102N21A:IXYS)
4.Faster Pulse Trans.s
(Finemet:hitachi)
New Thyratron gate board
1.Opt. link 30ns
2.FET Driver 38ns
3.MOS-FET 19ns
①
②
4.Pulse Trans. 23ns
① Delay time from ext. trigger to optical link input
55ns
② Delay time from ext. trigger to grid pulse
165ns
②-① Delay time from optical link input to grid pulse 110ns
Results of Delay Time for Kicker P.S.
600
[ns]
500
400
Target
Line(<290ns)
300
200
5
10 15 20
35 [kV]
Technical Problem 1: Shortening of delay time form trigger
Further improving plan
1. Further shortening of delay time in FET-Drivers
by replacement them.
(38ns→25ns)
2. Further shortening of delay time in pulse transformers by reducing leakage inductance.
(23ns→16ns)
3. Deleting transition time of optical link to change
a high voltage PFL circuit.
(30ns→0ns)
Target of a total reduction delay time is 50ns.
Delay time target
275ns → 225ns
Plan of changing the PFL circuit
Z0=25 Td=120nZ0=25 Td=25n Z0=25 Td=120n
R1
SW1
100kV
L1
25
T1
T2
T3
V1
R2
Optical Trigger
(30ns delay)
Z0=25 Td=120n
R1
100kV
V1
T1
D2
Z0=25 Td=25n Z0=25 Td=120n
D1
SW1
Electric Trigger
(no delay)
T2
L1
T3
25
R2
Technical Problem 2: shortening of charging time
Z1
600V:100kV
10ms,0.04A
600V
V1
D1
Z0=25 Td=120ns
T1
K1
Z1
100μs,4A
600V:90kV
600V
V1
D1
K1
Main Charge
500kHz
M1
100μs,0.4A
M3
600V:100kV
Z0=25 Td=120ns
T1
Hybrid
Charging
System
D2 D4
Sub Charge
M2
M4
K2
D3 D5
Technical Problem 2: shortening of charging time
Main Charge
Sub Charge
100%
Voltage
90%
Hybrid
Charging
System
4Ap
0.4Ap
500kHz
0
100μs
time
200μs
summary
 We developed the fast response kicker system.
 Faster optical link, FET driver, MOS-FET and pulse
transformer were adopted in the new thyratron gating
board, and the obtained response time of the kicker
power supply was 275ns. It was short for realization
of the self-trigger individual injection method with an
energy of 200 MeV/nucleon.
 In order to get further margin of the response time,
we will try to improve by (1)replacement of FET driver,
(2)reduction of leakage inductance of pulse
transformer and (3)changing the PFN circuit.
 Our kicker system will be used not only for injection
but also for extraction. So high-speed repetition
charging system is necessary. We have begun to
examine the feasibility of a hybrid charging system.
Thank you very much!
Create New World with Pulsed Power