Transcript BARNES_CTF3_kicker_CollaborationMeeting_2008

CTF3: KICKERS & PULSERS
AT CERN
M.J. BARNES & T. FOWLER
Michael Barnes
CERN
AB/BT
Also representing:
Tony Fowler, Gianfranco Ravida
& Hiromitsu Nakajima
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg1
M.J. BARNES & T. FOWLER
CTF3 Programme
2005
2004
Thermionic gun
Linac
DL
CLEX 2007-2009
building in 2006
Photo injector / laser
tests from 2006 2007
30 GHz production
(PETS line)
and test stand
Kickers for CTF3
Extraction
kicker
Tail
clipper
: kicker
M.J. Barnes
TL2 2007
CR
2006
2007
X2; 2007
operation
CTF3 Collaboration Meeting, 22/01/2008
pg2
M.J. BARNES & T. FOWLER
CTF3 & CLIC Stripline Kickers
CTF3 CR
Extraction
CLIC CR
Tail
Clipper
300
2500
200
MeV
Total kick deflection angle (“B” & “E” Fields)
7
2.5
1.2
mrad
Stripline plate separation
40
20
40
mm
Stripline length
1.7
3
Beam energy
Available length (including transitions)
m
1.625
m
Field rise & fall-times (0.25% to 99.75%)
≤ 70
≤ 30
≤5!
ns
Pulse duration
200
50 to 60
Up to 140
ns
Flat-top reproducibility
± 0.1
± 0.1
NA
%
Flat-top stability (including droop)
± 0.25
± 0.25
NA
%
± 15
%
5
Hz
Field homogeneity
Initial
5
Nominal
50
150
50
Hz
Pulse voltage
12.6
10.5
2.4 for 1m
kV
Pulse current (into 50 Ω load)
252
210
48
A
≤ 1 rms
ns
Repetition rate
Timing Jitter
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg3
M.J. BARNES & T. FOWLER
Combiner Ring Extraction Kicker
For operation in 2008 the existing magnetic kicker, which only partially satisfies the
specifications, will be replaced by a new stripline kicker designed and built by
CIEMAT.
Tests were undertaken at INFN
Frascati:
Vacuum achieved 2e-7 mbar with
turbo pump and no bake out;
HV pulse test at 16 kV, 5 ns pulse
width. HV DC test at 18 kV;
RF measurements (0 to 1 GHz)
compare well with HFSS
simulations;
Conical transition pieces at the
extremities are designed to reduce
impedance discontinuities;
Recently delivered to CERN.
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg4
M.J. BARNES & T. FOWLER
Combiner Ring Extraction Kicker
Stripline kicker powering
Generator
#1
 The electrodes must be pulsed with separate, opposing
polarity power supplies. Physically located in Bât. 2002.
+ve
Generator
#2
−ve
 These pulsed power supplies have been be obtained by
modification of ex-EPA equipment and change of system
impedance from 30 Ω to 50 Ω. They have been pulse
tested up to 17.5 kV, 200 ns into dummy 30 Ω loads.
 Positive pulse generator uses ex-EPA INJ GEN4.
Electronics are in racks RA042 and RA043.
 Negative pulse generator uses ex-EPA INJ GEN3.
Electronics are in racks RA040 and RA041.
HFSS
3D Model of strip-lines
Beam
with rectangular aperture
 Controls will use existing DSC (Device Stub Control)
interface (DCTFPOW1, rack RA110) presently used for
magnetic extraction kicker control.
 Installation during February/March 2008.
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg5
M.J. BARNES & T. FOWLER
Tail Clipper: Overview
Beam Pulse
≤5ns
The beam pulse extracted
from the CR is 35 A and 140
ns. The tail-clipper must
have a fast field rise-time, of
5 ns or less, to minimize
uncontrolled beam loss. The
flatness of the kick pulse is
not important as deflected
beam is to be thrown away.
KICK
Beam
Pulse
from
CR
35A
0A
140ns
140ns
Schematic Of Tail Clipper
PFN
HVDC
Fast
switch
Z
Transmission
line
Z
Copt
From second
pulse generator
Stripline plates
Z
Beam
Z
Termination
resistor
Z
Z
Each pulse generator is composed of a 50 Ω Pulse Forming Network (PFN), a fast
semiconductor switch, 50 Ω stripline plates and a matched terminating resistor.
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg6
M.J. BARNES & T. FOWLER
Tail Clipper: Rise Time
• To make use of both the electrical and magnetic fields to deflect the beam, the
striplines must be “charged” from the CLEX (beam exit) end;
• To provide 1.2 mrad deflection, with striplines terminated in 50 Ω, requires 2.4kV
on 1 m long striplines;
• Electrical pulse propagation through 1 m striplines, at speed of light, takes 3.3 ns!
• Tail Clipper requires short field rise-time (≤ 5 ns) !.
• In order that stripline fill time does not significantly effect deflection rise-time,
several sets of striplines, mechanically in series, will be used.
For a trapezoidal current pulse:
L 

Start pulse at  T0 

Nc


Start pulse at T
2L
T  5ns  Tr 
Nc
0
Where:
To
CLEX
Beam
Bunch
L/2
M.J. Barnes
L/2
Kickers for CTF3
Tr is permissible pulse rise time;
L is overall length of strip-lines;
N is number of sections;
c = 3 x 108 m/s
L  1 m & N  2 , Tr =1.7 ns
For:
L  1 m & N  3 , Tr =2.8 ns
CTF3 Collaboration Meeting, 22/01/2008
pg7
M.J. BARNES & T. FOWLER
Tail Clipper: Semiconductor Switch
Options considered for fast semiconductor switch
≈
Stacked MOSFET Switches (SMS)
have been developed at TRIUMF:
these systems have voltage ratings
up to 12.5 kV and have been tested
at burst-rates of up to 3 MHz. Recent measurements on an SMS gave
10% to 90% rise-time of 3.7 ns for
a 4.8 kV pulse.
M.J. Barnes
Behlke, high voltage, MOSFET
switches are available with suitable
datasheet rise-times. 5 kV (HTS-5008-UF) and 8 kV (HTS-80-20-UF)
versions of the switch have been
tested.
Kickers for CTF3
It is planned to evaluate a Fast
Ionization Dynistor (FID), the FPG
5-01M122S144N, which has a
specified output voltage of 5 kV
into 50 Ω with a rise-time of 1.5 ns
to 2 ns.
Order Placed Mid-June 2007:
Not Yet Received
CTF3 Collaboration Meeting, 22/01/2008
pg8
M.J. BARNES & T. FOWLER
Tail Clipper: SMS Update
For an SMS device the relative timing of the turn-on of each MOSFET is
critical: thus it is important that the trigger to each MOSFET is coherent
in time. Sample Fibre Optic transceivers, the SLM-001-IS-PTN from DLightsys, have been extensively tested.
500MHz scope
100MHz probe
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg9
M.J. BARNES & T. FOWLER
Tail Clipper: SMS Update 2
CERN Specifications for Fibre-Optics:
• The fibre-optic receiver is required to output a signal which can vary in duration from DC to a pulse train
consisting of 50 ns wide pulses with a 50% duty-cycle (i.e. 10 million pulses per second for CLIC Combiner
Ring Extraction Kicker). Within these lower and upper frequency bands the pulse train can have an arbitrary
duty-cycle which is defined by the input signal;
• The timing of an edge on the output of any receiver should be within 1 ns of the timing of the same
edge from any other fiber optic transmitter-receiver pair.
Measurements on SLM-001-IS-PTN (5pps, Vcc=5.5V):
255.9
DLY Time [ns]
80
81.6
250
70
240
60
230
50
220
40
210
30
200
24.1
20
190
187.3
10
180
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15
Excel row number
M.J. Barnes
Kickers for CTF3
• Spread in front edge delay with
respect to trigger ≈ 58ns (1ns
specified) !!;
260
WID(C2) Time [ns]
Front edge DLY(C1, C2) [ns] min
Front edge DLY(C1, C2) [ns] MAX
Back edge DLY(C1, C2) [ns]
WID(C2) [ns] min
WID(C2) [ns] MAX
90
• Spread in width with respect to
trigger ≈ 69ns (1ns specified) !!;
• Spread in back edge delay with
respect to trigger ≈ 5.2ns (1ns
specified).
CONCLUSION
SLM-001-IS-PTN transceivers are
NOT suitable for SMS, and thus a
suitable fibre optic has NOT yet been
identified.
CTF3 Collaboration Meeting, 22/01/2008
pg10
M.J. BARNES & T. FOWLER
Tail Clipper: Behlke Switch Update
Transmission: 10m of
50Ω coax (HTC-50-7-2)
PFN: 20m low-loss 50Ω
coax (HTC-50-7-2)
Load Resistor (RT818A, 50Ω
nominal) & fast CT (CT-B1.0B)
Tektronix TDS5054B
HVDC Power Supplies
Low Inductance connections
to Behlke switch
Screw connections to Behlke
Behlke HTS-80-12-UF switch
2 x 10pF (Copt)
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg11
Tail Clipper: Behlke Switch
Measured Current
M.J. BARNES & T. FOWLER
Switch: Behlke HTS-80-12-UF (S/N 745136) with 2 x 10pF (Copt); −5.7kV PFN
voltage; 50Hz operation. [500MHz scope; 500MHz current-transformer].
FAST (3ns rise) trigger input.
Results:
• 200 ns flattop;
• −59 A load current (−48A
required);
• 2.352 ns Max fall (10%
to 90%) [7.6x106 pulses],
(permissible: 2.8 ns for 3
section stripline !);
Trigger: 1V/div
• Time-jitter of load current
w.r.t. trigger: σ=36.41 ps
(3σ = 0.11 ns);
• Trigger jitter …..
Load Current: 10A/div
M.J. Barnes
Note: RT818A load resistor (2W)
temporarily reduced from ~50Ω to
~44Ω due to temperature rise.
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg12
Tail Clipper: Behlke Switch
Summary of Measurements
M.J. BARNES & T. FOWLER
Summary of Test Results for Behlke Switch Type HTS-80-20-UF
Permissible value of current Rise/Fall  ~2.80 ns for a 1 m, 3 section stripline
Behlke
Switch
S/N
Behlke
Connector
Pulse
Generator
10%→90%
rise (ns)
PFN
Voltage
(V)
Rise/Fall
(10% to
90%) (ns)
Time-jitter
w.r.t.
trigger [σ]
(ps)
3σ+
trigger
jitter..?
(ns)
Number
of
“Statistic”
Pulses
640242
Push-Pin
3
+5700
2.233 
114.7
0.34+?
1.96x106
640242
Push-Pin
3
+6050
2.285 
232.5
0.70+?
7.91x106
640243
Push-Pin
3
+6050
2.257 
115.4
0.35+?
4.22x106
745136
Screws
3
−5700
2.352 
36.41
0.11+?
7.60x106
745136
Screws
3
+5700
2.378 
66.73
0.20+?
0.44x106
745136
Screws
31
− 5700
2.401 
131.9
0.40+?
0.99x103
Load solder joint deteriorating (sparking)
To achieve fast rise/fall of field and low time-jitter:
• Integrity of connections is very important;
• Low-loss coaxial cables are required;
• A “fast” (e.g. 3 ns rise-time) trigger pulse, with low time and amplitude jitter, is required.
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg13
Tail Clipper: Behlke Switch –
Predicted Field Rise-Time
M.J. BARNES & T. FOWLER
Measured 48A scaled; 3x0.33m striplines
Measured current: 3x0.33m striplines
Measured 48A current; 4x0.33m striplines
Measured current: 4x0.25m striplines
Stripline Maximum Field (%)
M.J. Barnes
Kickers for CTF3
105%
3.5ns
120
110
100
90
80
70
60
50
40
30
20
10
0
-10
4.6ns
Measured current: 3 series 1.1ns striplines
50.6A
V(StripLine_In)/50
V(StripLine_Out)/50
StripLine Field
-0.8%
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
5
4.9
4.8
4.7
4.6
4.5
4.4
4.3
4.2
4.1
4
3.9
100
102
104
106
108
110
112
114
116
118
120
122
124
126
128
130
132
134
0% to 100% Field Rise Time (ns)
3
1st
trough
<100%
(0.33m)
Current (A) and Field (%)
2
50R
1
Time (ns)
• For an overall stripline length of 1 m, a current of
~48A provides the specified kick of 1.2 mrad;
• Measured current used to predict field in (ideal)
striplines  3 series 0.33m plates meets 5ns rise-time;
• Predicted field rise, 0% to 100%, can be reduced
from ~4.8ns to ~4.0ns by using 4 series 0.33m plates
(however 1625mm overall length must be respected);
• A 4th set of striplines also provides redundancy;
• “Over-driving” striplines also reduces field rise-time.
CTF3 Collaboration Meeting, 22/01/2008
pg14
Tail Clipper: Behlke Switch
Summary
M.J. BARNES & T. FOWLER
• For
an overall stripline length of 1 m, a current of ~48 A is required to achieve the
specified kick of 1.2 mrad.
• The on-state resistance of the HTS-80-20-UF Behlke switch, when conducting ~60A
of load current, is approximately 9 Ω.
• With a 50 Ω PFN and 50 Ω load, a PFN voltage of 5.25 kV is required to achieve
48A of load current. The HTS-80-20-UF has been tested at up to 6.05 kV PFN
voltage at 50 Hz: however, for long-term reliability, it is desirable to operate at < 70%
of 8 kV switch rating (5.6 kV).
• A load current rise-time of ≤ 2.4 ns, 10% to 90%, is consistently achieved with the
HTS-80-20-UF Behlke switch and low-loss transmission line.
• Predicted field rise-time, 0% to 100%, can be reduced from ~4.8ns to ~4.0ns by
using 4 series 0.33m striplines instead of 3 series 0.33m (however 1625mm overall
length must be respected). The 4th set of striplines also provides redundancy allowing
operation with nominal kick with increased rise-time.
• Length of each set of striplines will be chosen:
• to use 4 series striplines;
• to maximize overall length of striplines (to optimize deflection versus switch
stress).
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg15
Tail Clipper: Behlke Switch
Proposed Installation
16 Cables
Entrance
KHC 0250
M.J. BARNES & T. FOWLER
Striplines x4
Collimator &
internal dump
Striplines x4
Enclosure for
Behlke switches,
PFNs, terminators,
scope & chair.
Bât. 2009
Enclosure
SKETCH OF CTF3 TRANSFER LINE WITH KICKER BEHIND WALL ….
• Behlke switch, control electronics and load are all outside high radiation environment;
• Low-loss coaxial cable (HTC-50-7-2) will be used for PFN (~16 m), transmission cable from
switch to striplines (~10 m) and for transmission cable from striplines to load (~10 m).
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg16
M.J. BARNES & T. FOWLER
Conclusions
• CIEMAT have designed, constructed and tested 50 Ω striplines for extraction from the
CTF3 CR. Initial RF tests show good agreement with design simulations. The device is
now awaiting vacuum tests before installation in the machine.
Tail-Clipper:
• SLM-001-IS-PTN transceivers (D-Lightsys) are NOT suitable for a SMS system; the
spread in delays is excessive. Thus a suitable fibre optic has NOT yet been identified.
Therefore SMS technology is not being pursued for tail-clipper. But for CLIC CR…...
• For a 1 m stripline length, subdivided into 3 sections, the predicted 0% to 100% field
rise-time, with current measured for a Behlke HTS-80-20-UF switch, is ~4.8ns.
• Predicted field rise-time, 0% to 100%, with the Behlke HTS-80-20-UF, can be
reduced from ~4.8ns to ~4.0ns by using 4 series 0.33m striplines. The 4th set of
striplines also provides redundancy allowing operation with nominal kick but with the
increased rise-time.
• The Behlke HTS-80-20-UF switch provides adequate performance: an order for these
switches will be placed shortly (3 month delivery time; system installation in 09/2008).
• A Fast Ionization Dynistor, the FPG5-01M122S144N, has not yet been received; this
FID will be tested but will not be used for the tail-clipper.
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg17
M.J. BARNES & T. FOWLER
Questions ??
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg18
Behlke HTS-80-20-UF: Measured
Current
M.J. BARNES & T. FOWLER
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
-0.1
50A (norm)
60A (norm)
Δtrise (10%→90%): ~0.1ns
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
11.5
12.0
12.5
13.0
Normalized Current
Behlke HTS-80-20-UF: Measured Current
Time (ns)
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg19
Striplines: Angular Deflection Due
To Magnetic and Electric Fields
M.J. BARNES & T. FOWLER


V *l *c

 E  arctan 
9
 d * ( p *10 ) *  * c 


 0.3   V 
 *l
 B    * 
 p    d  * c 
 2  


These equations show that the deflection due to the
magnetic field is independent of the impedance of the
striplines.
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg20
Agilent HFBR-2528 Fiber Optic:
Receiver Delay
M.J. BARNES & T. FOWLER
Delay of Edge (s)
Delay of Edges through HFBR-2528 Receiver (various pulse widths, positive,
TTL input pulse) with a Reference HFBR-1528 Transmitter
1.20E-07
1.19E-07
1.18E-07
1.17E-07
1.16E-07
1.15E-07
1.14E-07
1.13E-07
1.12E-07
1.11E-07
1.10E-07
1.09E-07
1.08E-07
1.07E-07
1.06E-07
1.05E-07
1.04E-07
1.03E-07
1.02E-07
1.01E-07
1.00E-07
9.90E-08
9.80E-08
Delay(Ch2,Ch1) - back edge - 200ns
Delay(Ch2,Ch1) - back edge - 5us
Delay(Ch2,Ch1) - back edge - 100ns
Delay(Ch2,Ch1) - front edge - 200ns
Delay(Ch2,Ch1) - front edge - 5us
Delay(Ch2,Ch1) - front edge - 100ns
>15 ns spread in delay
0
M.J. Barnes
20
40
60
Kickers for CTF3
80
100
120
140
160
180
200
Excel Row Number
CTF3 Collaboration Meeting, 22/01/2008
pg21
Agilent HFBR-1528 Fiber Optic:
Transmitter Delay
M.J. BARNES & T. FOWLER
1.110E-07
1.105E-07
1.100E-07
1.095E-07
1.090E-07
1.085E-07
1.080E-07
1.075E-07
1.070E-07
1.065E-07
1.060E-07
1.055E-07
1.050E-07
1.045E-07
1.040E-07
1.035E-07
1.030E-07
1.025E-07
1.020E-07
1.015E-07
1.010E-07
1.005E-07
1.000E-07
9.950E-08
9.900E-08
9.850E-08
9.800E-08
~8 ns
~5 ns
Delay(Ch2,Ch1) - back edges - 200ns
Delay(Ch2,Ch1) - front edges - 200ns
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
92
96
100
104
108
112
116
120
124
128
132
136
140
144
148
152
156
160
164
168
172
176
180
184
188
192
196
Delay
back
edge
(s)(s)
edges
Delay -of
HFBR-1528 Transmitter Delay, measured data sorted on back edge delay,
with a "Reference" HFBR-2528 Receiver
Arbritary Number
M.J. Barnes
Kickers for CTF3
CTF3 Collaboration Meeting, 22/01/2008
pg22