T18_VG2.4_Disk_#2 processing
summary
SLAC Workshop
090708-10
T. Higo and Nextef group
Whole history of T18_#2 processing
•
•
•
•
•
•
•
•
090709
Installation in Oct. 2008
Steady-state run at 80MV/m in Dec. 2008
Higher field in Jan-Feb 2009
Breakdown rate meas. in Mar.-Apr. 2009
Longer pulse in latter half of Apr. 2009
Breakdown rate meas. May 2009
Various measurement in June 2009
Finish with 4000 hrs operation
SLAC Workshop
2
Processing history for 9 months
Nextef startup and careful processing
T18_VG2.4_Disk #2 Processing whole RF-ON history
4000
3500
RF-ON vs date
(hr)
3000
2500
2000
1500
1000
090709
SLAC Workshop
2009/7
2009/6
2009/5
2009/4
2009/3
2009/2
2009/1
2008/12
2008/11
0
2008/10
500
3
Recovery pass in (T,P) space
P
SLAC
P
CERN
P
KEK
50%P, -20%T
20%P, 20nsT
T
T
aP, -1stepT
T
In case of number of BD’s even at
lower power level than defined
090709
SLAC Workshop
4
Whole BD history of T18_VG2.4_Disk #2
090610
MasterTable_Eacc_Trend till_090610
120
3000
100
2500
80
2000
1500
60
<Eacc> 51nsec Ushi
<Eacc> 113ns Ushi
<Eacc> 173n ushi
<Eacc> 213n Ushi
<Eacc> 253ns Ushimoto
Eacc keep 253ns Ushi
Eacc 51ns Ushi
Eacc 113ns Ushi
Eacc 173ns Ushi
Eacc 213ns Ushi
Eacc 253ns Ushi
Eacc Usahimoto 2/21-3/23
Eacc 252ns Usahi 4/1-4/7
Eacc 252ns 4/7-14
Eacc 312ns 4/14-16
Eacc 412ns 4/23-27
Eacc 331ns 4/28-5/1
Eacc 252ns 5/1-11
Eacc 412ns 5/11-6/10
40
20
0
0
090709
Total# BD
<Eacc> MV/m
Total BD
500
1000
1500
2000
2500
3000
RF-ON SLAC Workshop
integrated
(hr)
3500
1000
500
0
4000
5
Nextef: RF monitors along waveguide
Low loss waveguide
2-klystrons  120MW, 300ns
klystron—3dB 2.3m
3dB— LLWG 1.5m
LLWG 5m
LLWG—ACC Str 5.0m
WR90 flanges ~36sets*7mm SUS
ACC Structure
090709
Loss Kly.Str. ~23%
Delay ~ 50ns
SLAC Workshop
Two klystrons
6
Nextef: Monitors along beam axis
PM GV
ACC str
GV
GV FC
GV
KX03
VAC
Q-mass
VAC DC
FC-UP Qmass
090709
Slit
AM
1 87
PM
Varian
IP
VAC
230
WC
Insul
425
PM
FC
DC
E
FC
Lounine
IP H load
IP
Varian IP
H IP
PM
AM
E
DC
WC
FC
IP H Otsuka
load
Load
FC-Mid
SLAC Workshop
FC-DN
7
Now T18_VG2.4_Disk #2 has been
processed since late October
GV
FC
090709
SLAC
3dB hybrid
Acoustic Plastic scintilator
sensors
& PMT
T18_VG2.4_Disk
SLAC Workshop
(These are not used effecively.)
8
Big breakdown followed by a breakdown
from the very first pulse
file number= 3 20081124_210211_1 Red=final, black=previous pulse and blue=-2nd pulse
30 000
30 000
24 000
DPO1
25 000
Fs
15 000
10 000
Ra
20 000
20 000
10 000
5000
10 000
200
18 000
200
200
400
600
800
200
1000
30 000
30 000
20 000
20 000
10 000
10 000
400
600
400
600
800
1000
400
600
800
1000
10 000
800
1000
Tr
5000
5000
0
DPO2
Rs
22 000
20 000
Nominal BD
10 000
30 000
30 000
20 000
Fs
20 000
10 000
10 000
200
400
600
10 000
800
1000
FC-UP
20 000
30 000
200
400
600
10 000
800
1000
200
FC-Mid
20 000
30 000
200
400
600
800
1000
10 000
800
1000
20 000
1st pulse BD,
following big BD
30 000
10 000
20 000
5000
24 000
25 000
20 000
22 000
15 000
20 000
10 000
200
10 000
18 000
200
200
400
600
800
200
1000
400
600
800
1000
30 000
20 000
20 000
10 000
10 000
400
600
800
1000
5000
5000
0
30 000
600
30 000
file number= 4 20081124_210301_1 Red=final, black=previous pulse and blue=-2nd pulse
30 000
400
10 000
400
600
800
1000
10 000
10 000
30 000
30 000
20 000
20 000
10 000
10 000
200
400
600
800
1000
200
10 000
10 000
20 000
20 000
30 000
30 000
090709
400
600
800
1000
200
400
600
800
1000
10 000
200
10 000
SLAC Workshop
400
600
800
1000
20 000
30 000
9
Current burst toward upstream
30 000
20 000
10 000
20 000
22 000
Fs
15 000
Ra
30 000
Rs
24 000
25 000
20 000
5000
10 000
200
10 000
18 000
5000
200
0
200
400
600
800
200
1000
30 000
30 000
20 000
20 000
10 000
10 000
400
400
600
800
1000
400
600
800
1000
10 000
600
800
1000
Tr
5000
10 000
30 000
30 000
Fs
20 000
20 000
10 000
10 000
200
10 000
20 000
30 000
400
600
800
1000
FC-UP
200
10 000
20 000
30 000
400
600
800
1000
200
400
600
800
1000
10 000
FC-Mid
200
10 000
400
600
800
1000
20 000
30 000
Rare event but need to understand the mechanism
One out of 100 breakdowns
Abrupt big burst in current only to upstream
no change in RF pulse shape
090709
SLAC Workshop
10
Power at BD’s and BDR during 27 days at 80MV/m (35MW)
during RF-ON from 700 ~ 1200 hours since startup
Started at
10:00 on
BD power vs RF-ON period
Nov. 28
Finished at
9:00 on
Dec. 25
from Nov. 28-Dec. 25
Total 20 BD's/585hrs (3 BD's are over pulse driven)
Power
MW (calib. needed)
50
566 hrs run / 643 hrs
= 88% ON
40
30
BDR
20
~17 str. BD / 566 hr
10
~0.030 BD/hr/str
Power
MW
Excessive power related
Power FC-UP
Power is calculated from DPO with a background subtraction
0
0
090709
100
200
300
400
RF ON Period (hours)
SLAC Workshop
500
~0.75 x10-6 BD/Pulse/m
600
11
252ns, 65MW
Total operation hours >5MW = 140.6 hours
Period ACC-IN > 64 MW = 91.7 hours
090709
SLAC Workshop
12
Appearance of breakdowns at 65MW run
Breakdowns and other INTLK's during 65MW run
Run#29
09027-090305
2070
2060
Total BD
2050
Total BD
ACC-BD
Following BD
FC-UP
WG-BD
2040
2030
2020
2010
2200
2250
2300
2350
RF-ON (hrs)
090709
SLAC Workshop
13
252ns, 70MW
090709
SLAC Workshop
14
Appearance of breakdowns at 70MW run
Breakdowns and other INTLK's during 70MW run
Run#32-2
090403-090407
2160
Total BD
2155
Total BD
ACC-BD
Following BD
FC-UP
WG-BD
2150
2145
2140
2720
2730
2740
2750
2760
2770
2780
2790
2800
RF-ON (hrs)
090709
SLAC Workshop
15
252ns, 75MW
090709
SLAC Workshop
16
Appearance of breakdowns at 75MW run
Breakdowns and other INTLK's during 75MW run (Run#32-3)
2300
2280
Total BD
2260
Total BD
ACC-BD
Following BD
FC-UP
WG
2240
2220
2200
2180
2850
090709
2860
2870
2880
2890
RF-ON (hrs)
SLAC Workshop
2900
2910
2920
17
Typical processing / steady-state run
• Breakdowns are
– mostly followed by a few to several successive
breakdowns
– typically from the very first pulse at even lower
power.
• Spurious events exist
– with flush of current towards upstream occur
from time to time
090709
SLAC Workshop
18
Breakdown rate evaluation
T18_VG2.4_Disk #2
Breakdown rate for 252ns and 412ns
BDR(ACC)
[1/pulse/m]
10-4
0423-0427
(2900hr)
BDR(ACC) 412ns
BDR(ACC) 252ns
BDR(ACC) 252ns
0411-0414
(2800hr)
0227-0305
(2300hr)
0402-0403 (2700hr)
10-5
0520-0525
(3500hr)
0525-0530
(3500hr)
0403-0407 (2700hr)
0401-0402 (2700hr)
0313-0323 (2500hr)
-6
10
0530-0610
(3700hr)
1128-1225
(700-1200hr) 0511-0515
(3300hr)
0515-0518
(3400hr)
0501-0507
(3100hr)
10-7
80
85
90
95
100
105
110
115
120
Eacc [MV/m]
090709
It is not so clear from this experiment that the breakdown rate
decreases as function of “processing.”
SLAC Workshop
19
Breakdown rate versus Eacc
Breakdown rate for 252ns and 412ns
BDR(ACC)
[1/pulse/m]
10-4
BDR(ACC) 412ns
BDR(ACC) 252ns
BDR(ACC) 252ns
10-5
10-6
y = 2.3888e-18 * e^(0.28603x)
R= 0.97016
y = 1.815e-17 * e^(0.24289x)
R= 0.94792
10-7
80
85
90
95
100
105
110
115
120
Eacc [MV/m]
Exponential fit.
090709
SLAC Workshop
20
Breakdown rate for 252ns and 412ns
BDR(ACC)
[1/pulse/m]
10-4
BDR(ACC) 412ns
BDR(ACC) 252ns
BDR(ACC) 252ns
10-5
10-6
y = 5.2958e-61 * x^(27.542)
R= 0.96504
y = 2.1053e-60 * x^(26.721)
R= 0.94346
10-7
80
85
90
95
100
105
110
115
120
Eacc [MV/m]
Cannot discuss functional form due to large scatter.
090709
SLAC Workshop
Power fit.
21
R. Ruth at X-band workshop at CI, UK, Dec. 2009
RF Processing of the T18 Structure
RF BKD Rate Gradient Dependence for 230ns Pulse at Different
Conditioning Time
RF BKD Rate Pulse Width Dependence at Different Conditioning
Time
-4
10
-4
10
After 500hrs RF
Condition
-5
10
After 900hrs RF
Condition
-6
10
After 1200hrs RF
Condition
-7
10
95
100
105
110
Unloaded Gradient: MV/m
G=108MV/m
BKD Rate: 1/pulse/m
BKD Rate: 1/pulse/m
After 250hrs RF
Condition
10
10
G=108MV/m
-6
G=110MV/m
10
115
-5
-7
100
150
200
RF Flat Top Pulse Width: ns
This performance maybe good enough for 100MV/m structure for a warm collider, however, it does not yet
contain all necessary features such as wake field damping. Future traveling wave structure designs will also
have better efficiencies
7/2/2016
X-Band 2008
Page 22
Breakdown rate versus width
BDR_vs_Width
BDR [1/pulse/m]
10-4
10-5
10-6
BDR = 3e-18 * Tp^5
BDR = 9e-25 *
Tp^7.45
-7
10
0
100
200
300
400
500
Width [nsec]
Not enough to discuss functional form.
But it is evident that the longer pulse makes breakdown rate
large with much more than linear dependence.
090709
SLAC Workshop
23
Summary of breakdown rate
• Exponential slope as Eacc
– Data scattering band around a slope is by an order
of magnitude
– Slope = an order of magnitude by 10 MeV/m
– Not enough to discuss about the functional form
• Breakdown rate evolution
– Reduction of BDR was not seen in last 2000 hours
090709
SLAC Workshop
24
Breakdown position and timing
• Still trying to analyze in detail.
• At present, mainly refer to Steffen’s previous
analysis
– More frequent at downstream
090709
SLAC Workshop
25
Analysis still on the way
137 examples of run24: 60MW, 400ns
Rs_timing
DPO run34
Rs-Tr
Rs_rise
60MW, 400ns
Tr_rise
350
500
300
400
250
300
200
200
150
100
100
run34_60MW_400ns
timing delay from Rs to Tr
100
50
Rs-Tr
600
Rs_rise, Tr_fall
Rs, Tr Timing and Rise/Fall time
Tr_timing
0
-50
-100
50
0
-100
0
20
40
60
80
100
120
0
140
-150
0
20
40
60
80
100
120
140
BD*#1
BD*#
Timing of Rs rise and Tr fall show correlation.
Two cases: those at the same time and those at 50ns later.
Position and time of the breakdown should be deduced from Rs and Tr.
FC current burst timing and amount should be integrated in the analysis.
Breakdowns from the very first pulse are missing; should be included in the analysis.
090709
SLAC Workshop
26
Steffen 090227
Breakdown position for 205 ns data
16
14
No. of shots
12
10
8
6
4
2
0
5
10
Breakdown position
15
Red real cell timing, blue linear cell timing, 205 ns data
Dark current measurement
• Setup
FC-UP
FC-Mid
75cm
75cm
AM
FC-DN
195cm
• Amount and beta value of dark current versus
processing time
• Spectrum at RF-ON 2000hrs and 4000hrs
090709
SLAC Workshop
28
Amount of dark current
Dark Current 090414 Tp=252ns
100
FC-Mid = 30mA
@100MV/m
1
FC-UP ~ FC-Mid/3
FC-UP
microA
10
0.1
FC-DN ~ FC-Mid/4
FC-UP microA (253ns, 090225)
FC-Mid microA (253ns, 090225)
FC-DN microA (253ns, 090225)
0.01
60
70
80
90
100
110
120
Eacc [MV/m]
090709
SLAC Workshop
29
T18_#2 Dark Current evolution
081128-081224-090224-090414-090515
100
100
10
10
microA
T18_#2 Dark Current evolution
081128-081224-090224-090414-090515
1
FC-UP [microA] (081128)
FC-UP [microA] (081224)
0.1
1
FC-Mid
FC-UP microA
Dark current evolution 252nsec
FC-UP
microA
(253ns, 090225)
FC-UP
microA
(253ns, 090414)
FC-Mid [microA] (081128)
FC-Mid [microA] (081224)
0.1
FC-Mid
microA
(253ns, 090225)
FC-Mid
microA
(253ns, 090414)
FC-Mid
microA
090515
FC-UP microA 090515
0.01
50
60
70 80 90 100
Eacc [MV/m]
200
0.01
50
60
70 80 90 100
200
Eacc [MV/m]
Measured at RF ON 700 – 1200 – 2100 – 3000 – 3400 hours
Decreased by a factor 2 between processing with max Eacc 80  110MV/m,
but no more suppression in the following steady-state run for more than 1000 hours
090709
SLAC Workshop
30
Dark current evolution 252nsec
T18_#2 Dark Current evolution
081128-081224-090224-090414-090515
T18_#2 Dark Current evolution
081128-081224-090224-090414-090515
10-3
10-3
I_UP/Eacc^2.5
(081128)
FN FC-UP (253ns, 090414)
10-4
I_UP/E^2.5
253ns
090515
beta=36
beta=37
10-5
10-6
0.008
microA/(MV/m)^2.5
FN FC-UP (253ns, 090225)
y = 0.14557 * e^(-653.42x)
R= 0.99838
y = 0.074674 * e^(-672.19x)
R= 0.99772
0.01
0.012
0.014
0.016
(081128)
I_Mid/Eacc^2.5
(081224)
FN FC-Mid (253ns, 090225)
FC-Mid
FC-UP microA/(MV/m)^2.5
I_UP/Eacc^2.5 (081224)
I_Mid/Eacc^2.5
FN FC-Mid (253ns, 090414)
10-4
I_Mid/E^2.5
253ns
090515
beta=40
10-5
beta=39
y = 0.22345 * e^(-596.01x)
R= 0.99982
y = 0.1351 * e^(-607.18x)
R= 0.99854
-6
0.018
10
0.008
0.01
0.012
0.014
0.016
0.018
Eacc [MV/m]
Eacc [MV/m]
Measured at RF ON 700 – 1200 – 2100 – 3000 – 3400 hours
No big change in shape nor slope (beta).
090709
SLAC Workshop
31
Deduction of the field enhancement factor
Fitting of modified F-N curve
I
e
2.5
E
6.53109  1.5

 E s (V / m )
e

a
Eacc ( MV / m )
Assuming Es/Eacc=2
actually T18_VG2.4_Disk
Es/<Eacc>~2.62 max
(Cu)=4.52eV
6530  1.5
23951


a ( Es / Eacc )
a
090709
SLAC Workshop
32
Width dependence
Darck current at 103MV/m
dependence on pulse width
1000
FC-UP [pA]
FC [pA] at 50Hz
800
Upstream current
behaves as linear on
width
FC-Mid [pA]
600
Downstream current has
two components.
400
200
0
0
100
200
300
400
090515
Width
090709
500
SLAC Workshop
33
Dark current spectra in June
Dependence on power
T18_VG2.4_Disk #2
Spectrum_vs_Width at 64MW
090618
T18_VG2.4_Disk #2
Spectrum_vs_power at 252ns
1.2
I_slit_peak [microA]
1.4
1
I_slit_peak [microA] 64MW, 331ns
I_slit peak [microA]
I_slit_peak [microA] 55MW, 252ns
I_slit_peak [microA] 70MW, 252ns
I_slit peak [microA]
64MW, 252ns
I_slit_peak [microA] 113ns, 64MW
I_slit_peak [microA] 64MW, 252ns
1.2
09
Dependence on width
1
0.8
0.6
0.4
0.8
0.6
0.4
0.2
0.2
0
0
0
0
5
10
15
5
10
15
pc (MeV/c)
pc (MeV/c)
Actual field of analyzer magnet was checked.
The formula used up to now pc[MeV/m] = 1.646 x I [A] = 8.23xRef. Volt. [V] was confirmed.
Two peaks appear and higher for higher momentum one.
Less than ½ of full acceleration.
Little exists below 2.5MeV/m.
090709
SLAC Workshop
34
Dark current versus operation frequency
090618
T18_VG2.4_Disk #2
Downstrean dark current vs frequency
350
FC-UP [pA]
FC-DN [pA]
FC-UP, FC-DN [pA]
300
250
200
150
100
11422.5
11423
11423.5
11424
11424.5
11425
11425.5
11426
Freq [MHz]
Increase as operation frequency increases.
Easy to be captured for both!?
090709
SLAC Workshop
35
Dark current behaviour
• Amount
– to upstream << to downstream
– Divergent: 1/4 from Mid to DN
– Linear vs pulse width
• Reduction of dark current
– By a factor 2~3 during first 2000 hrs
– No reduction after 2000 hrs
• Beta value
– Stayed almost constant from 700hrs to 3300 hrs
– Not processed?
• Spectrum
– Two peaks below half of full acceleration
090709
SLAC Workshop
36
Conclusion
• Established a basic procedure of processing and evaluation
– Can proceed a series of structure tests
– Better to further develop a system in such as
• Phase measurement, missing energy evaluation, etc.
• Breakdown rate was evaluated.
– Gross comparison with that of SLAC is consistent with each other.
– In order to precisely compare, power estimation and identification of
breakdowns should be better.
• Dark current
– Decreased during initial processing but stayed constant during a longterm run without higher field nor longer pulse processing.
– Energy is at most a half of full acceleration and little exists below a few
MeV/c.
• Breakdown pulse analysis
– Breakdown pulses are still to be analyzed carefully.
090709
SLAC Workshop
37