Highlights of the YAG measurements carried out October 9-13, 2007

Highlights of the YAG Highlights of the YAG measurements carried out measurements carried out

October 9-13, 2007October 9-13, 2007

Recycler Weekly Departmental MeetingOctober 10th, 2007

L. Prost, A. Shemyakin, A. Warner,M. Sutherland, K. Carlson, A. Burov

2

Thanks to all involved !Thanks to all involved !

Instrumentation setup including installing the camera in the tunnel and proper timing in the control room took significant efforts from Arden, Mary, Kermit, Randy

Additional problem with SPAGNI failure and Kautz power station outage Great responsiveness from EE and Mechanical support

• Thanks to Ron, Jerry, Fernando, Rich, Jeff, Greg, Kermit

Taking and analyzing the data was a collective effort from Mary, Arden, Kermit, Timofei and Alexey (& Sasha and I)

A lot had to happen in a short period of time to make these measurements fruitful They were very successful

• Hopefully, it will directly translate into faster cooling !

3

GOALGOAL

Before the shutdown, YAG images were taken and showed that the beam at the exit of the cooling section was very elliptical, indicating a strong quadrupole component Large angles

The goal of the measurements was to correct the ellipticity Make the beam cylindrical in the cooling section i.e.

round AND low envelope scalloping

4

Camera setup (I)Camera setup (I)

Camera located at the exit of the cooling section (past the 180 bend magnet)

Electron beam direction

CID camera

180 bend magnet

YAG crystal drive (towards the outside wall)

Note:The constant focusing direction is y (vertical) for the camera but actually x (horizontal) for the beam.

y

x

SPQ01I (not visible on the picture) is the last solenoid before the YAG

Beam reference system

5

Camera setup (II)Camera setup (II)

Camera settings Screen = 6 kV MCP = -600 V Gate/Photo cathode = -200 V Gate width = 100 ns

• Gate position:~Max of the beam pulse

YAG crystal

Ti wires for calibration

6

TimingTiming

Ref. timing for the gate delay based onSPQ01I = 0 A image Drift of the order of 2 s was observed over a shift

• Corrected byadjusting gatedelay

Acquisition window

Gate pulse

Control electrode pulse

‘Real’ gate position w.r.t. beam pulse depends on cable length…

Ref. image (inverted)

Adjusted timing to return to this image

7

In the y-direction (constant focusing) From geometrical considerations: 15 pixels/mm Confirmed with measurements with inserted scraper and

YAG motion• Images of a wire placed in front of the YAG crystal overlap

when the YAG is moved by 3 mm and the corresponding image shifted by 45 pixels in the YAG direction of motion (x-direction)

Trials to do direct measurements for calibration in the other (x) direction have so far failed !! Data are inconsistent

CalibrationCalibration

Sum of two images:YAG at position 1 and YAG at position 2, with the image shifted by 45 pixels

Image of the wire in front of the YAG

x

y

8

Distortion determination (I)Distortion determination (I)

Ideally, the YAG crystal makes a 45 angle with respect to the axis of the camera Gives distortion k ≡ cotan between X and Y directions

At small currents, an effect of a solenoid is reduced to a beam turn by .Due to a possible deviation of the YAG tilt from 45 degrees, the visible angle is different from its ideal value , which is described as

At << 1, it is equivalent to

)2/( BBl

)2cos()1(1/ kdd

dI

dB

B

lCkCdId

2;)2cos()1(1/

Use SPQ01I to verify/determine k

A. Burov

In the YAG setup, this is the X-direction of the camera which is affected by this angle

cotan

9

Put scrapers in to create a ‘line of reference’ which will rotate as a function of SPQ01I

Identify rotation of the whole beam by fitting an ellipse

Distortion determination (II)Distortion determination (II)

y = 2.3628x + 51.036

R2 = 0.9992

45

47

49

51

53

55

57

59

61

63

65

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

SPQ01I [A]

An

gle

of

the

fitt

ed e

llip

se [

deg

]Fitted ellipse

angle

Threshold image (auto)

Slope is d/dI

10

Using

and taking two scraper data sets, we find C = 2.4/A (design: 2.3/A) and k = 0.84 (i.e. 50 angle instead of 45)

Additional data sets (i.e. SPQ01 scans and associated rotation of the fitted ellipse) agree rather well

Distortion determination (III)Distortion determination (III)

)2cos()1(1/ kCdId

0

0.5

1

1.5

2

2.5

3

0 20 40 60 80 100 120 140 160 180

Initial angle [deg]

d

/dI [

de

g/A

]?

d/dI calculated with C = 2.4/A and k = 0.84

Dependence of d/dI on the apparent angle

11

Beam rounding procedure (A. Burov, Dpt Meeting July 25Beam rounding procedure (A. Burov, Dpt Meeting July 25thth))

For two distinct values of SPQ01I Record initial image Change upstream quads successively (6 quads)

• Record associated images

Calculate ellipticities Fit ellipse to threshold (binary) image Extract semi-major and semi-minor e = 2 (a –b)/(a+b) Include effect of camera angle (i.e. distortion k)

Compute MULT SVD algorithm

Use MULT to make the beam more round Repeat…

• If e = 0 for two values of SPQ01I, then the beam is perfectly round

This was done automatically (eventually) with a Java application written by Timofei Boshakov

12

Note on the image analysisNote on the image analysis Once the image is recorded, the first step of the analysis

consists in creating a binary image, for which a threshold (0-255) is set, somewhat arbitrarily This threshold determines how much of the beam (‘core’,

‘core + halo’,…) is taken into account in the calculation of the best fit ellipse

• For the automation process we choose a threshold value of 55 i.e. selection of the core of the beam

Original image (inverted)

Threshold

Fit ellipses

Auto (here 22) 33 (i.e. ‘core’) 4 (i.e. ‘halo’)

13

Preliminary results (Preliminary results (after multiple iterationsafter multiple iterations))

First visible image + 1.5 s

118 119 120

121122123

Nominal

New nominal

Quads off Quads on SPB01I SPB02I SPQ01I Area Major Minor Angle Ellipticity118 X 19.1 8.7 0 5812 116.067 63.757 2.155 0.37119 X 19.1 8.7 10.4 1655 63.519 33.174 21.537 0.52120 X 13.5 12.1 10.4 913 42.654 27.253 32.664 0.40121 X 13.5 12.1 10.4 1000 38.682 32.916 144.048 0.20122 X 19.1 8.7 10.4 1941 52.483 47.089 163.214 0.10123 X 19.1 8.7 0 6871 101.519 86.176 37.57 0.19

14

Some untold details…Some untold details…

Once the beam for our nominal settings was considered round enough Switched to DC mode and adjusted correctors such as

to recirculate 200 mA Verified that steering did not introduce different

quadrupole components• Made an additional scan with quadrupole

– Ellipticities found were similar to before steering• We did see smaller ellipticities

than reported in previous table– Timing drift ?

Checked pepper-pot image• Fairly rectangular and regular

pattern

15

Adjusting ‘Adjusting ‘cylindricitycylindricity’’

Additional SPQ01I scans are needed to determine /correct for the magnitude of envelope scalloping The scan gives information about the beam radius,

angular momentum and radial divergence (fit) Back propagate to the entrance of B1 lens in OptiM

• Get initial conditions Adjust B1 and B2 to have a cylindrical beam in the CS

Applied this procedure to our nominal file (Alexey) Got new settings forSPB01I & SPB02I

• High current values obtained maybe a problem (see further)

Repeated rounding procedure withnew settings

• Analysis of the final resultsunderway

YAG screen

16

Higher order multipoles seem to be generated in the first Higher order multipoles seem to be generated in the first bendbend

Upstream (of the first bend) focusing and steering scans show very large effects on the beam shape while downstream scans primarily change the beam size and rotation. Example of a scan with SPA06 (last solenoid before the first

bend) at 190G in the cooling section (and corresponding field at the cathode)

• This effect likely more pronounced at 190G because of the larger effective emittance (thus beam size)

SPA06I = 14.4 A SPA06I = 15.4 A SPA06I = 16.4 A SPA06I = 17.4 A

Nominal

17

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0 0.2 0.4 0.6 0.8 1

Beam current [A]

Elli

pti

cit

y

Some (possible) consequenceSome (possible) consequence

Because the beam size increases for higher current (it actually has a maximum at ~250 mA), the multipole component of the beam may be differentat different beam currents Example for the nominal case ‘rounded’

• Not obvious… More visible on a non-optimal

case

Upulse = 3.2 kVIeq =0.1 A

Upulse = 4 kV

Ieq = 0.3 A

Upulse = 4.5 kVIeq = 0.6 A

Ellipticity increases ~2×

18

Quality of the solenoidal field appears to degrade at high Quality of the solenoidal field appears to degrade at high currentscurrents

When the lens (SPQ01 or SPB01 for instance) current becomes relatively large (1/2 of its maximum?) the beam shape starts degrading (i.e. does not remain round or at least elliptical) High non-linearities at high field, even near the axis ?

SPQ01I scan, Upulse = 4.5 kV

19

‘‘Last minute’ analyses (Last minute’ analyses (from Alexeyfrom Alexey))

Fit of SPQ01A scan for the ‘new’ nominal file: Give r = 2.4 mm, = 0.01, = 2.0 m for initial conditions

at the entrance of the cooling section Show that 10-15% envelope oscillation remain

• Improvements expected with SPB01I = 20.4 A & SPB02I = 8.0 A

22.28380

Wed Oct 17 11:40:10 2007 OptiM - MAIN: - Y:\MI-31\OptiM Files\BackFromQ1ToB1.opt

0.3

0

90

-90

Be

tatr

on

siz

e X

&Y

[cm

]

An

gle

[de

g][

-90

,+9

0]

a b Angle[deg]

5 10 15 20

0.125

0.15

0.175

0.2

0.225

0.25

0.275

SPQ01I [A]

Bea

m r

adiu

s [c

m]

SPQ01 scanw/ fit

Corresponding envelope in cooling section

~0.6 mmr = 2.4 mm

20

ConclusionConclusion

Beam rounding converged and we achieved ellipticities of the order of 10-20% Two files available:

• Before shutdown nominal corrected for its ellipticity• Improved nominal (new SPB01 and SPB02 settings

intended to make the beam more cylindrical in the cooling section)

– SPQ01 scan indicates that ~15% oscillation may remain (other settings proposed by Alexey could be tried)

Did one set of rounding measurements at 190 G• Not as successful as for 105 G

– We seem to have too strong sextupole components

Several questions remain Consistent calibration of the x-direction Effect of steering/focusing upstream of the first bend Possible non-linearities from lenses at high current

• i.e. any change to lenses with already relatively high current changes the multipole component in the beam