• Aucun résultat trouvé

Experimental determination of impedance and delay time of the 100 $\Omega$ meander transmission line for the SPIRAL2 Single Bunch Selector

N/A
N/A
Protected

Academic year: 2021

Partager "Experimental determination of impedance and delay time of the 100 $\Omega$ meander transmission line for the SPIRAL2 Single Bunch Selector"

Copied!
4
0
0

Texte intégral

(1)

HAL Id: in2p3-00620766

http://hal.in2p3.fr/in2p3-00620766

Submitted on 27 Oct 2011

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Experimental determination of impedance and delay

time of the 100 Ω meander transmission line for the

SPIRAL2 Single Bunch Selector

M. Di Giacomo, P. Balleyguier, Ac. Caruso, F. Consoli

To cite this version:

M. Di Giacomo, P. Balleyguier, Ac. Caruso, F. Consoli. Experimental determination of impedance and delay time of the 100 Ω meander transmission line for the SPIRAL2 Single Bunch Selector. 2nd International Particle Accelerator Conference (IPAC2011), Sep 2011, San Sebastian, Spain. TUPS075, pp.1710-1712, 2011. �in2p3-00620766�

(2)

_ * # † C

EXPERI

OF

Abstract

A 100 Oh selector of th non standard pulser powe measurement used to mea electrodes. To reduce selector [1] static magne dump, and t along high selecting the impedance Z critical issue (=v/c)= 0.0 The first s was measu oscilloscope detail in [2]. also perform the meande density the st Figure According manufactured with differen number of m the available Table Parameter Meander len Al203 thickn Meander num ___________________________ *Work supported #digiacomo@ganNew affiliation: CP 65-00044 Fra

IMENTA

F THE 10

P

F. Co

hm meander l he SPIRAL2 d d characteristic er but introdu ts. The paper asure the imp

INTRO

e the requir of the SPIRA etic field defle

two short pul impedance e bunch to be

Zc and the del

es of the devi 4 beam transp set of electrod ured with a and gave Zc= Power tests med and a man r u-bends, l trip can handl

e 1a : former (a g to the concl d with modifi nt dimension meanders was space along t e 1: difference gth (mm) ness (mm) mber _________________ d by EU commissi il.fr Associazione scati, Rome, Italy

AL DETER

00 OHM M

SPIRAL

M. Di G

P. Balleygui

onsoli

, A. C

ine is used in driver medium c impedance h uces the prob

describes the pedance and

ODUCTION

red power, t AL2 driver a ecting the bea

lses of oppos (100 Ω) me e accelerated. lay τ of the e ice, located a port line. des that have b an impedanc =94 , =0.04 with a contin nufacturing pro limiting the le. a) a) and present lusion of [2], ied shape as sh ns as reported modified due the accelerator e in meander p Set 1 46 3 45 ion 7th framework Euratom/ENEA y

RMINAT

MEANDE

L2 SINGL

Giacomo

#

, G

ier, CEA/DA

Caruso, A. L

n the single b m energy line helps to reduc blem of calib e different me the delay o

N

the single b accelerator u am towards a site sign trave eander electr The characte electrodes are along the in a been manufac ce-meter and 44, as describ nuous current oblem appeare maximal cu t (b) u-bend a second set hown in Fig. d in table 1. e to a reducti r. parameters Set 2 50 4.3 39 k project n. 21269 A sulla Fusi

TION OF

ER TRAN

LE BUNC

GANIL/Spi

AM/DIF,

F-Longhitano

bunch e. The ce the brated thods of the bunch ses a beam elling rodes, eristic e then a beta ctured d an bed in were ed on urrent b) t was 1 and The ion in Se chara elect intro The chap

Refl

To a m begin as sh Figu equiv Zi know wher impe prop refle obtai S11 chara 50 Ω have stand mean mm. in ha the s 92. ione,

IMPEDA

NSMISSI

CH SELE

iral2, Caen,

-91297, Arp

o, INFN-LN

everal techni acteristic par trodes with oduction of m methods and pters.

MEAS

flection Mea

o measure refl matching resis nning and at t hown in Fig. 2

ure 2: the elect valent circuit in can be calc wn transforma

=

re Z and Z edances of pagation const ection coeffic ined from: is measured w acteristic imp Ω. The model e to be meas dard 50 Ω ada nder involved Zml and kml h aving Rl=50 Ω second port of Zcn2 Lcn2 kcn2 Zin

ANCE AN

ON LINE

ECTOR

*

, France

pajon, Franc

NS, Catania,

iques were rameters Zc a a network matching netw the results ar

UREMENT

asurements

flection at the tor (Rl) and

the end of the 2. trode measure for simulation culated by ap ation equation

are the ch each sectio tants, and Li cient and the

=

with the netwo pedance of th l parameters sured aside o apters and con d in the beta have to be det Ω to use either f the network a Zml Lml kml

ND DELA

E FOR TH

ce

Italy

used to m and  of the k analyser work, not easy e reported in

T METHO

input of the m connexions e meander line ement assemb n model (b). pplying iterati n [3]: haracteristic a on, γ = α is the related e S11 param ork analyser a he analyzer p of the conne or are already nnectors. The of the line, i termined. We r calibrated co analyser. Zcn1 Lcn1 kcn1

AY TIME

HE

measure the e new set of without the y to calibrate. the following

ODS

meander plate (cnx) at the e are required a) b) bly (a) and the

vely the well (1) and the load ik are the d lengths. The meter can be (2) and Z is the port, equal to cting devices y known for e longitudinal s Lml = ~273 are interested oaxial loads or Rl e f e . g e e d e l ) d e e e ) e o s r l 3 d r

TUPS075 Proceedings of IPAC2011, San Sebastián, Spain

1710 Copyright c○ 2011 by IP A C’11/EPS-A G — cc Cr eati v e Commons Attrib ution 3.0 (CC BY 3.0) 07 Accelerator Technology T30 Subsystems, Technology and Components, Other

(3)

Case 1 : 50 Ω Load.

Figure 3: reflection measurements principle with

unmatched ending load.

The network port measures the load impedance transformed by the meander line. Due to the effect of the transformation, points of minimum and maximum of Zin

are repeated periodically as shown in Fig. 4, where the magnitude of S11 and the polar plot of the reflection

coefficient are represented.

From the measured plots and considering the lowest frequency values to minimize parasitic effect we obtain:

 S11max = -3.61 dB or Zmax= 246.4 Ω at 11.48 MHz

 S11min = -24.49 dB or Zmin= 55.5 Ω

 Δf = 22.96 MHz

Zml can be calculated from two formulas:

Zmax=Zml2/Zmin (3)

|

| ≈

(4)

In the last case only the Zmax value is used.

Values of 116.8 Ω and 109 Ω have been obtained from the two methods, while 112 Ω is the value for which the analytical model fits better the measured results.

The frequency difference Δf between two minima is

the frequency at which the meander length Lml is

equivalent to λ/2. The electrode  can then be calculated from:

 = 2Lml f /c (5)

Case 2: Open and Short Circuit

Another classic method consists to end the unknown line with a short and open circuit and to measure the corresponding reflection patterns

open and

short shown

in Fig. 5.

The phase difference of the two signals gives a quasi sinusoidal curve (Fig. 5b) whose extremes correspond to frequencies for which the length of the line is equal to λ/8. The first value is 5.7 MHz, corresponding to  = 0.0415.

In this case, Zc is given by the formula [4] :

=

(6)

We used open and short circuits from a 50 Ω calibration kit, and the effect of the adapter is probably not negligible. The calculated impedance appears like in Fig. 6, and an average value around 115 Ω can be estimated at low frequency.

Figure 6: Estimation of Zc from the open/short method.

Case 3: Matched Line

If a variable termination is used in the scheme of Fig. 3, it is possible to reduce the circle to a quasi single point, centred at the value of the meander impedance as shown in Fig. 7a. Looking in detail, a small oscillation is still present (Fig. 7b) due to the reactive component of the trimmer. The measure of the trimmer resistance indicates 112 Ω but this value appears to be somehow smaller than the position of the circle centre on the diagram, situated around 115 Ω.

Figure 7: Reflection measurements pattern with rather matched ending load and corrected delay.

Transmission Measurements (S

21

)

The meander is loaded with a matching resistor and the S21 parameter is measured twice, for different positions of

port 2 (28 meanders) as shown in Fig. 8. Network Analyzer 1 50  load Meander line Impedance = 100 (a priori) Cable and network

Impedance = 50 n = 39 périodes 0 20 40 60 80 100 100 110 120 130 0 0 f (MHz) Zc() 1 0.5 0 0.5 1 1 0.5 0 0.5 1(a) 0 10 20 30 40 50 110 112 114 116 118 120 Zc() f(MHz) (b)

Figure 4: reflection measurement patterns with non matched ending load: S11 and normalised Z.

Figure 5: Reflection measurements pattern with closed and open circuits.

0 20 40 60 80 100 4 3 2 1 0 0 20 40 60 80 100 180 90 0 90 180 f (MHz) dB degré dB short open

(a) amplitude (b) phase

open

-180°

short

Proceedings of IPAC2011, San Sebastián, Spain TUPS075

07 Accelerator Technology

T30 Subsystems, Technology and Components, Other 1711 Copyright

c○ 2011 by IP A C’11/EPS-A G — cc Cr eati v e Commons Attrib ution 3.0 (CC BY 3.0)

(4)

Figure 8: Measurement of beta by phase difference. A 10 kΩ resistor is used to reduce the perturbation of the line impedance, nevertheless we observed that the matching resistor, placed at the end of the meander line, had to be changed slightly (6-7 Ω) between the two positions of port 2. From the difference between the phase delays (S21) at both positions, one can calculate the

delay of the considered number of meanders and the relative propagation speed (beta) of the electrodes:

 ()=(2-1)/;  = Lml /(c) (7)

Errors dues to port connections cancel with this method, while those due to impedance perturbation stay. We couldn’t cancel the residual ripple visible in Fig. 9 and due to residual mismatching between the line and the ending load.

Figure 9: examples of phase measurement and calculated delay.

MEASUREMENT RESULTS

Results from the different method are summarised in Table 2 where S11 and S12 indicate a reflection or

transmission measurement.

Table 2 : results with different methods

Method Zc Beta S11 - 50 Ω load 117 0.0417 S11 - meas/model fitting 112 0.042 S11 - 50 Ω NA port2 109 0.042 S11 - open/short 115 0.0415 S11 - minimum reflection 114 S12 – phase difference 0.043

Expected figures were Zc=100 Ω and β=0.04.

Measured values are higher than expected (10 to 15% for Zc, a factor 2 less for the ) and their spread is not negligible. A parameter that could explain the difference is the relative permittivity of the Alumina which should

be around 8.4 instead of 9.9. First measurements seem to confirm this hypothesis.

Concerning the spread, a more detailed analysis is in progress to take off the effect of the connections and to identify key points in the measurement procedure.

Other Measurements

We have also measured the spread among different plates and the sensitivity to the screw tightening. The four units show very similar results, confirming the reproducibility when built all together. Concerning the screws, it’s paramount to tighten those at the extremities and in the middle of the plate at least, as shown in Fig.10.

Figure 10: Sensitivity of the amplitude (dB) of the maximum and the minimum of the S11 curve of Fig.4a

and of the frequency (MHz) of the minimum.

CONCLUSION

Several methods have been tested to determine the electrical parameters of non standard, high impedance travelling wave electrodes. Results still present some spread but clearly indicate higher Zc and delay than expected from the Mafia and CST Microwave Studio simulations. The difference could come from the permittivity of the Alumina.

REFERENCES

[1] G. Le Dem, High Frequency Chopper - final report, EMDS report I-020274, GANIL, May 2009.

[2] P. Balleyguier, M. Di Giacomo, G. Fremont,

M. Michel, P. Bertrand, “Electrode design

improvements in the SPIRAL2 single bunch selector", LINAC2010, Tsukuba, Japan, 12-17 September 2010.

[3] D. M. Pozar - Microwave Engineering, Third Edition, Wiley, 2004.

[4] W. Johnson, "Transmission Lines and Networks", McGraw-Hill, 1963, p.155.

[5] P. Balleyguier, F. Consoli, M. Di Giacomo "Mesures du deuxieme prototype d’antenne à meandres", EMDS report I-027517, GANIL-SPIRAL2, July 2011.

[6] F. Consoli, P. Balleyguier et M. Di Giacomo "Analysis of the measurements performed on March 2011 on the Single Bunch Selector electrodes”, EMDS report I-027517, GANIL-SPIRAL2, July 2011. Network analyser 1 2 10k 10k R=Zc n= 28 pe riods 0 20 40 60 80 100 -180° -90° 0° 90° 180° f (MHz) 1 2 0 20 40 60 80 100 14 14.5 15 15.5 16 f (MHz)  (ns) 23 23,1 23,2 23,3 23,4 23,5 23,6 23,7 23,8 -40 -35 -30 -25 -20 -15 -10 -5 0 0 2 at I/O 2+2 center 2+2+4 2+2+4+8 0 MHz dB Number of screws min max freq

TUPS075 Proceedings of IPAC2011, San Sebastián, Spain

1712 Copyright c○ 2011 by IP A C’11/EPS-A G — cc Cr eati v e Commons Attrib ution 3.0 (CC BY 3.0) 07 Accelerator Technology T30 Subsystems, Technology and Components, Other

Figure

Figure 7: Reflection measurements pattern with rather  matched ending load and corrected delay
Figure 10: Sensitivity of the amplitude (dB) of the  maximum and the minimum of the S 11  curve of Fig.4a  and of the frequency (MHz) of the minimum

Références

Documents relatifs

clairement ce qui rapproche ces femmes dans une seule catégorie pour la communauté migrante : c’est le fait, qu’à la différence des autres migrantes sénégalaises, épouses « à

We allow users to deploy innovative experiments across our federated platforms that include the embedded object testbeds of FIT IoT-Lab, the cognitive radio testbed of FIT

Changing Failure Rates, Changing Costs: Choosing the Right Maintenance Policy..

comparison with the ground range planimetric distance ρ (Fig. 4: Illustration of relief effects on the ground coverage of the radar beam. 5: Definition of the parameters used in

The authors gratefully acknowledge Louis Collins and Alan Evans for providing the SPAM data, Brian Scanlan (Thieme) for giving permission to use the Talairach’s plates contained

Comme il a de plus été élaboré pour analyser les interactions entre les temps de travail et les temps de formation, il nous semble particulièrement pertinent pour mener une analyse

Malgré l’apport de la théorie de l’agence pour appréhender la composition du conseil d’administration dans les entreprises familiales, plusieurs failles

diteur, de collectionneur de livres d’artistes et d’animateur d’un réseau d’artistes enveloppant plusieurs continents, l’idée du livre d’artiste que Maurizio