Fandom

Scratchpad

ERLP DC Gun Commissioning Procedures

216,193pages on
this wiki
Add New Page
Discuss this page0 Share

Ad blocker interference detected!


Wikia is a free-to-use site that makes money from advertising. We have a modified experience for viewers using ad blockers

Wikia is not accessible if you’ve made further modifications. Remove the custom ad blocker rule(s) and the page will load as expected.

ERLP DC GUN COMMISSIONING: PROCEDURES v. 4.3 Yuri Saveliev

CONTENTS

HV DC gun conditioning 1 HV conditioning after reactivation 2 Photocathode QE 2 YAGs characterisation 3 Beam steering 3 BPM calibration for Q measure. 4 Dipole calibration 4 Energy spread 5 Beam halo 6 Buncher set-up 6 Bunch length 8 Transverse emittance 9 Photocathode lifetime 11 Off-axis beam generation 11 Injector optimisation 11 Bunch charge (N/A) 12



___________________________________________________________________


DC gun electrode conditioning Prerequisites: 1) the gun is baked-out and vacuum achieved \~10^-9^-10^-10^mbar. 2) leak detection carried out Hardware/DAS: 1) analogue signal of ion pump current continuously monitored - chart recorder.(the extractor gauge should not be operated at higher >300kV voltages) 3) PSU current monitor - 4) HV probe signal - to chart recorder 5) HV ripple signal - on scope 6) "fool-proof" HV limiter (so that not to exceed the HV pre-set level by accident) Agree but you also need current limiter. I am assuming you have a high impedance conditioning resistor installed which will severely limit the current. Otherwise arcs can do real damage. A slow automatic ramp up and down would be desirable now and necessary in the future. If you turn off HV you don\’t want to just slam it off. \[GN\] Yes, we do have current limiter in controls console that switches the primary PSU off if exceeded. HV ramping up and down is also available for normal operating regimes. But if the current is above the limit the HV slams down at once (with no ramping). \[YMS\] A control to turn off the HVPS off is good but not sufficient unless you are only driveing the voltage from a microamp-level supply . The problem is that the stored energy will still be sufficient to damage the cathode or the ball. Hence a high impedance resistor. Also a high volume arc can spray junk (molten metal chunks which then field emit permanently) onto the cathode. We learned this the hard way. \[GN\]

8) Is HV interlocked to gun gate valve (which I assume is interlocked to vacuum trip point)? Does HV turn off if gate valve is closed?. Need to have residual gas analyzer available if needed. Doesn\’t have to be on line all the time. \[GN\] It would be good to have a high sensitivity beam viewer flag to watch continuously also. Set solenoid to nominal calculated value. You can watch flashes as the system conditions. It would be nice to have TV camera viewing cathode during this procedure. If hot spots occur you may be able to see where it is happening. Also if you have an arc you will want to see if the cathode is damaged. Do I understand this is in the line of the drive laser beam dump? How are you going to avoid blinding the camera when you view the laser straight on? \[GN\] Unfortunately, the beamline installation/baking-out will be done in parallel with the gun electrode conditioning. Hence both having a YAG screen and viewing the cathode are not possible. Good points but … \[YMS\] NOTE: all controls/signals must be gathered in one place (manual controls acceptable). 2) x-ray detector (50-300keV range; as sensitive as possible) - N/A 7) some computer data storage/visualisation - requested, will be available - proved to be of limited use.

Procedure. Gradually increase voltage from \~50kV to \~400kV. Goes fast up to 150 kV then more slowly after that. You want to increase voltage initially in 5 kV steps and let it sit 5 min at each point. If you see emission spikes then let them condition away by sitting as long as it takes. If vacuum rises by > 5x then you must back off 5 to 10 kV and allow the vacuum to recover. Before continuing. It is possible to do this procedure automatically if you have a controller to run things. It looks for emission and if it sees none and vacuum is OK it steps up. \[GN\]


HV conditioning after cathode reactivation. Carlos H-G says: Once the electrodes have been conditioned as high as 370 kV, they remain conditioned up to that voltage. The idea behind conditioning the cathode rigth after activation is to condition just the cathode with its fresh layer of Cs on the surface! Cs lowers the work function down to 1.5 eV making any minute speckle more susceptible for field emission.

George mentioed that there was an arc that made a hole on the cathode, and before that there was observation of field emission current coming from the cathode. If the Cs channel spurious current was still present during the last activation, it is possible to add too much Cs on the cathode, leading to the excessive field emission. Do you know if the problem with the spurious Cs channle current has been fixed?

If you guys are confident that the electrodes have been conditioned to 370 kV, I would go ahead and condition the cathode right after activation. Here\’s how we condition our cathode:

1. Open the gun valve and make sure the solenoid is on and the YAG screen is visible. This will allow you to look at any field emission. 2. Increase the voltage from 0 to 50 kV, and soak there for abour 5 min. If you don;t see any field emission on the screen, you can soak for 2-3 min. If you see FE, then soak there longer until the FE vanishes or decreases. 3. Increase the voltage in steps of 50 kV and repeat observations mentioned in point 2. 4. Go up to 365 kV and soak there for about 20 minutes or unitl there is no FE seen on the screen. 5. ramp down to 350 kV and make sure there is no FE. 6. You are done.

Photocathode QE measurements The QE is measured according to: (for =532nm)

where Q is the bunch charge \[C\]; E~l~ is the laser pulse energy \[J\]; P~l~ is the laser power on the cathode \[W\]; I is the average beam current in the train \[A\], and  is the quantum efficiency.


YAG viewers characterisation Notes: - observation of the beam on YAGs is needed in the whole range of bunch charges up to a maximum of Q\~80pC. Hence some caution must be exercise not to burn the YAGs until we know the burn out threshold. - different conditions (full beam unfocused, full beam focused, slit image) will require different settings for the laser operation. Procedure: - set the laser PRF =5-10Hz (cameras are not synchronised, therefore not every "save" attempt is successful; having a higher PRF makes it less time consuming and less annoying). - set the laser power required and as low the laser pulse train length t as reasonable (maybe down to a single laser pulse in a train, i.e. t\~15ns ); - increase t until you are satisfied with the intensity of the image; - save the image and check the scans to ensure there is no saturation.

Procedure: "rule of thumb" for laser settings Aim: Determine typical laser settings for YAG imaging for different Q and conditions i.e. (i) full focused beam on "A"; and (ii) slit "B" and screen "C". - set Q (one in the range from 1pC to 80pc) and low t - increase t and grab images at each t; - determine X~rms~ each time and plot as function of t for a given Q; - hence determine "optimal" t for a given Q and condition (full beam or slit).

Beam steering

Notes. 1. Procedure for the beam steering in the injector line for an injector test must take into account an additional azimuthal rotation of the bunch from the coils inside the solenoids and Earth magnetic field. 2. Procedure for the beam steering in the injector line for a "full" ERLP operation, i.e. with the test line disassembled, should be developed. 4. x-y coupling in correctors inside the solenoids ! 4. How do we know that the beam is on axis at the entrance to the booster? (need for the pop-in viewer in \[A\] to be perfectly centred with respect to the BL axis and CCD camera must "see" not just the beam spot but the edges of the screen to make sense of beam position relative to the axis). - we\’ll have a second BPM in from of the booster entrance. This can also be determined empirically. Field level in cavity should not induce steering when beam is centered. \[GN\] Here I presume you are talking about the linac cavity. I guess the induced steering will also happen if the beam is injected into the booster at an angle to the axis. (Quite possible given the effect of Earth magnetic field I do not have however a feeling of how sensitive is the induced steering with respect to the initial transverse displacement/angle. Any comments? ). \[YMS\] It can certainly be calculated but we think we see effects at the level of order 1 mm or mrad. In our case there are asymmetric kicks due to the HOM couplers which we have to live with. I think your system will be better. Centering in solenoids can be done better than that. \[GN\]

Procedure (Test beamline): 01. Set U~DC~ and Q. (The solenoids settings hence the steering will depend on Q, therefore we cannot use lower than required Qs for steering). 02. Set PRF and train length (reduce train length to avoid saturation if necessary). 03. Set B~1~ and B~2~ (note : steering must be redone for any new solenoid fields settings )

1. Vary H&V-01 to centre the beam at BPM-01. 2. Vary H&V-06 to centre the beam at YAG-01 in "A". 3. Vary H&V-02 to centre the beam on YAG "B" 4. Vary field of SOL-02 and observe that the beam does not move on YAG "B". Otherwise, repeat steps 1-3 making small adjustments. 5. Check that the beam is centered on screens in "B" and "C" ensuring that the beam is parallel to the beamline axis.

Procedure (Final ERLP layout): 01. Set U~DC~ and Q. 02. Set PRF and train length (reduce train length to avoid saturation if necessary). 03. Set B~1~ and B~2~ (note : steering must be redone for any new solenoid fields settings )

1. Vary H&V-01 to centre the beam at BPM-01. 2. Vary H&V-06 to centre the beam at YAG-01 in "A". 3. Vary H&V-02 to centre the beam at BPM-02. 4. Vary field of SOL-02 and observe that the beam does not move at BPM-02. Otherwise, repeat steps 1-3 making small adjustments. 5. Vary Booster accelerating field an observe that the beam image does not move on YAG-02 (first YAG after the booster).

BPM calibration for bunch charge measurements Notes: - the BPM will be the only device to measure the bunch charge before the booster in the final ERLP layout. Procedure: This will need a linear op-amp (available - Alex).

Dipole calibration Notes: - HV PS output voltage used as a cross-check value for the dipole simulation data and for measuring the calibration curve of the dipole.

- For the dipole calibrations, use as low Q as possible to reduce the beam image (slit) enlargement due to energy spread within the bunch. 

- demagnetisation of the dipole may be needed: we could probably do it in situ with the portable gaussmeter.

HV probe - dipole modelling cross check Insert V-slit "A" on BL axis. Vary U~DC~ and adjust dipole current such that the angle of rotation is 45^o^ (beam centroid at the centres of the YAG "A" and YAG "D"). Plot I~d~( ) - should be a straight line in accordance with

where B~1~(s) is the dipole field at 1 Ampere excitation current. Compare the integral with computer simulated and field map measured.

Calibration curve. Insert V-slit "A". At given U~DC~, adjust current I~d~ to have 45^o^ rotation angle. Vary U~DC~ around a chosen value (+/- 10keV) and plot position of the beam centroid as a function of U~DC~. Expected to be a straight line in accordance with :

(some second order deviations from the straight line may take place).

Energy spread: correlated & uncorrelated P~b~ - buncher RF power; B~d~ - dipole magnetic field

Procedure 1 This procedure aims to obtain correlated and uncorrelated energy spreads as independent figures. 1. Set: B~1~=nominal; buncher=OFF; B~2~=0. 2. Steer the beam to the axis at a position of the centre of the dipole (x=0, =0). Use screens A, B, C. 3. Insert V-slit "A". 4. Set buncher phase =0. Then set P~b~=0. Set B~d~=nominal (beam centroid should be in the centre of "D"). 5. Increase P~b~ and set P~b0~ at which the width of the image "D" is minimal. (uncorrelated energy spread). At this P~b0~, the correlated energy spread ~Ec~ (\~5keV) in the bunch is compensated. Assuming (most probably rightly) that no new correlated spread will appear on the way between the buncher and the screen "D", this provides the opportunity to measure the uncorrelated energy spread on "D" (there may be required to set the =0 at P~b0~ if found that the phase setting drifts with the change of RF power). 6. Evaluate ~Ec~ from the value of P~b0~. 7. Register image on "D" at P~b~=P~b0~ => ~E~. (uncorrelated energy spread) Here we may assume that the emittance related width of the image ~~ is small due to the fact that YAG is positioned in the image plane of the dipole (this also should be independent of the beam energy) 8. Register image on "D" at P~b~=0 => ~tot~(correlated + uncorrelated).

Procedure 2 - insert V-slit "A", switch the dipole on, set =0 and =180^o^ ; grab images (in both cases the image expansion due to the uncorrelated energy spread is identical but the one due to correlated spread - is of opposite signs); - use some mathematical "wizardry" to extract ~E~ and ~Ec~. Notes: The buncher power here must be set below the level when the longitudinal cross-over takes place at "D". Otherwise, this may complicate analysis or result in wrong measurements.

Beam halo Measure vacuum with and without laser on the cathode (I now think we can hardly anything from this). Also viewer image w/o laser and Faraday cup signal under same conditions. \[GN\] There may be some misunderstanding (on my part) here. I think of a halo as a result of the laser beam not being perfectly "apertured" (e.g. long low intensity tails in transverse distribution or some reflections). Your comment I think refers to dark current that is also an important contributing factor. Is it correct? \[YMS\] Yes there can be field emissionin addition to imperfect laser illumination (or scattering of the reflected light from the cathode back onto the cathode. There can also be halo generated in srf cavities as field emission. It is a strong function of gradient. \[GN\]

Buncher set-up and characterisation

Procedure: buncher set-up. 1) Set solenoids 1 and 2 (get minimal beam size on screens "A" and "B") 2) Steer the beam to the axis at the buncher midplane and with the smallest angle possible - check the beam is on axis by observing the spot on screen "A". - check the angle is nil and the beam is on axis by varying SOL-02 field and observing the spot on screen "B" or "C". Spot must not move. 3) Set buncher rf phase. A. - dipole –ON, V-slit "A" - IN, observe spot on screen "E". - vary phase, find maximum positive and negative transverse exertions; the correct phase –exactly in the middle (plot x() for better accuracy). Notes: (i) phase /2 corresponds to minimal deviation angle wrt the injector axis. This is to ensure that =0 is set and not =. (ii) it is necessary to reduce the RF to such level that V~0~\~20-30kV; otherwise the beam spot will shift beyond the YAG "E" diameter, (iii) check linearity of phase setting before applying this procedure. B. (alternative procedure) - dipole=OFF; kicker =ON, V-slit "A"= IN; observe spot in "B" - set the kicker phase near on crest (kicker will deflect the bunch as a whole) - vary buncher phase; the slit image in "B" has maximal deflection at +/- /2. correct buncher phase (=0) is in the middle of the two above. This neglects non-linearity of rf phase settings. Probably can get more accuracy by varying gradient at nominal zero crossing and observing centroid does not move. \[GN\] You\’re right! This should probably give better accuracy. I\’ll put it into procedure. \[YMS\]- see \#4 below. 4) Check the phase setting by varying the RF power and ensuring the beam centroid does not move (with the dipole or kicker switched ON). 5) Set buncher rf power to a level required. - dipole – OFF, kicker – ON at =0; observe spot on screen "B".

Procedure: checking independence of the =0 setting on Pb. Dipole –ON, V-slit "A" - IN, observe spot on screen "E". A. - find the phase "knob" setting for =0 at several P~b~ values; - check the "knob" settings do not change with P~b~. B. - set =0 at one value of P~b~. - vary P~b~ in full range of power from zero to MAX; the beam image on YAG "E" should not move.

Procedure: checking linearity of the buncher phase setting. Notes: the linearity is required for choosing the correct buncher phase. Dipole =ON, slit "A" =IN; observation on YAG "E" (50mm diameter).

- vary phase +/- 180^o^ - adjust RF power such that the above phase variation does not drive the slit image beyond the YAG E boundaries. - plot beam position on "E" v "phase setting knob" reading - fit with SIN function.

Procedure for determination of buncher V0(Pb) function 1. Set: B~1~=nominal; buncher=OFF; B~2~=0 (if feasible). 2. Steer the beam to the axis at a position of the centre of the buncher (x=0, =0). Use screens A, B, C. 3. Insert V-slit "A". 4. Set buncher phase =0. Then set P~b~=0. Set B~d~=nominal (beam centroid should be in the centre of "D"). 5. Set P~b~. Vary buncher phase +/- /2 and find maximum x~max~ deviation of the beam centroid on "D". (should be symmetric wrt the beamlet position at P~b~=0). This corresponds to =/2 and maximum energy gain (loss) by the bunch. In a hard edge model of the buncher, this is given by (\*) (The more accurate value could be deduced from ASTRA/GPT modelling. Note: if  ; note that at 350keV we have 0.51 that still allows to approximate SIN with linear function with \~5% accuracy). 6. Scan a range of P~b~ and plot x~max~(P~b~) and corresponding V(P~b~) using dispersion function of the dipole. Compare with buncher modelling results.

Procedure for setting the buncher phase in the final ERLP layout. Use: BPM in front of the booster, fast scope, 1.3GHz reference signal. Principle: if the buncher phase is not at zero crossing, the bunch acquires (or loses) net energy that affects the time of flight between the buncher and the downstream BPM and, correspondingly, the phase of the BPM signal wrt the reference (1.3GHz) signal. This phase difference is estimated as:

where L\~0.8m is the buncher-BPM distance, =0.231m (for 1.3GHz), T-kinetic energy. If T=350keV, and T=10keV, we may expect \~10^o^ that should be noticeable. Procedure: 1. Steer the beam as normal. 2. Set P~b~=0 and note BPM phase relative to 1.3GHz reference signal. 3. Set the required P~b~ and adjust the buncher phase such that the relative phase stays the same as in Step 2. Additional possibility: if you monitor the power requested from the rf source to hold the gradient constant then during the bunch pulse the requested power should not change if the bunch is at zero crossing. \[GN\] What I mean is that our rf control system tries to maintain a constant field in the cavity by monitoring a probe in the cavity. If the gradient drops the rf systems drives harder. If you operate in a macropulse mode – say 250 us pulses at 2 Hz then you see on a scope the dip or rise in the power request signal (or equivalently the probe but that is a low level signal in the control loop so we don\’t like to mess with it) BTW this is also a good way to roughly phase the linac cavities initially when you have no idea of where the accelerating phase is. \[GN\]

Alternative procedure for setting the buncher phase in the final ERLP layout. The procedure is similar to that using the dipole magnet described above but here we use the steering coil (horizontal) located in the solenoid 2 and the BPM if front of the booster. Due to a small average angle of beam rotation, the expected effect is very small, e.g. less than 1mm at +/- 10keV. (???)

Bunch length measurements Notes. 1. Procedure of how to infer the actual bunch length at "B" from transverse spot measurements on screen "B" could be found in [\\apsv4\Astec\Projects\4gls\ERLP\Analysis-\\apsv4\Astec\Projects\4gls\ERLP\Analysis] ____________ 2. Apparently, the BPM can also be used for the bunch length measurements (by registering the signal in frequency domain). The BPMs after the light box and at the entrance to the booster could be used (even in a normal ERLP operation, i.e. with kicker removed). But their calibration is needed !. - NO WAY !!

Procedure Use: transverse kicker, V slit in "A" to aperture the beam, viewer in "B" for kicker RF phase setting and image grabbing. SOL-02 MUST be switched off when actual bunch length measurements are made! Measure the bunch charge on FC-01. Make sure the updated images of YAG screens in "A" and "B" are available (or store new ones if not certain). 1) Steer the beam to the kicker geometrical centre and with the angle wrt the axis as minimal as possible. (P~k~, P~b~=OFF). YAGs in "A" and "B". Vary SOL-02 field - ensure no beam spot movement on "B" then switch SOL-02 off. 1a) Acquire beam image on YAG "A". 2) Insert V-slit "A" at the beam centre. 3) Acquire slit image on "B": slit image position x~0~ and ~0~ (slit image RMS size on "B" without kicker). 4) Apply P~k~\~200-300W. Observe YAG "B". Change kicker RF phase  +/- 180^o^ and find  at which negative and positive deviations from axis on viewer "B" are maximal. Register x~max~ The phase =0 is in the middle. (Position should be the same x~0~ as above). Notes: (i) expected width of the slit image at =0 is \~10mm at full P~k~ . (ii) The image at x~max~ should be a thin line \~1mm wide and x~max~ values should be symmetric wrt x~0~. (In fact, the width will be probably dominated by emittance and is expected to be a few mm wide). (ii) linearity of kicker phase setting must be verified prior measurements in a similar fashion as with the buncher. 5) Set =0, check that the spot move symmetrically with small +/- phase variation. 6) Register (scan) the image on YAG "B". Longitudinal position z within the bunch wrt the beamlet centre relates to x-position on screen as

(here x and x~max~ are the values wrt x~0~). 7) Evaluate ~x~ of the image on "B" and calculate z~rms~=~z~ as follows:


Procedure Option 2. V-slit is not used - UNLIKELY TO GIVE ENOUGH ACCURACY ! The same as Option 1 but all transverse positions on screen should be taken as RMS values.


TRANSVERSE EMITTANCE MEASUREMENTS For validation of computer modelling, it is advisable to perform emittance measurements at both "A" and "B" locations. Also, at "B" - with and without the buncher.

Variables: 1) bunch charge Q 2) DC gun voltage U~DC~ 3) laser beam radius on cathode R~0~ 4) laser pulse duration (?) t 5) axial position of emittance measurements ("A" or "B") 6) buncher: ON or OFF 7) peak magnetic field in solenoids (B~1~ and B~2~ )

Potential complications: 1) background (slit edge scattered electrons); we may expect \~1-2% background \[12\]. If so, it can be neglected. 2) position and/or bunch charge instability during the emittance scans 3) Procedure I (two slits) is time consuming and moreover may prove to be difficult to implement due to small beam current value after two slits. Nevertheless, it should be attempted at least once for cross-checking the results. 4) Since the emittance is expected to be low, the beamlet width may not be "much larger" than the slit/pepper pot hole width. Correction may be needed.

Before starting emittance measurements (all procedures):

(i) Make sure the laser beam size and position on the cathode is known. Otherwise, store a laser beam image on the cathode and a reference photo of the cathode with LEDs on. (ii) Make note on magnet settings, DC voltage and other conditions that may likely affect the measurements and/or may be used later for cross-checks (e.g. laser system settings). (iii) Measure the bunch charge Q. (iv) Make sure the updated images of appropriate YAG screens are available (or store new ones if not certain). (v) Always ensure all images are indeed stored and are not saturated.


I. Two V slits ("A" and "B") Expected beam parameters (350keV) at "A": x~rms~= ~x~\~ 2-4mm; ~x~ \~ 0.8-1.5 m; divergence (correlated) < 5mrad Expected beamlet parameters at "B" (from V-slit in "A") Beamlet divergence (rms): \~ 0.2-0.8mrad; x~rms~\~ 0.2-0.8mm (at L\~1m)

1. Set Q, U~DC~, R~0~, B~1~. Set: Buncher=OFF, B2=0 (!!!) (demagnetise !). 2. Steer the beam to the centre of "A" and "B" using steerers 01 and 06 only (not 02) 3. Insert V-slit "A". Check the image in "B" is vertical. 4. V-slit "A" = OUT. Measure Q with FC 5. Store beam images on YAGs in "A" and "B" 5a. Store images of YAGs A and B with LEDs on (for image calibration purposes) 6. Scan V-slit "A" for a set of \{x~i~\}, i=10 and measure currents I~i~ with FC-01. 7. Scan V-slit "A" for the same set of \{x~i~\} as above and scan V-slit "B" over \{x~ij~\} sets (j\~10). Measure FC currents I~ij~. Check the background level. 8. Measure Q with FC (to compare with the initial value) 9. Store beam images on YAGs in "A" and "B" (to compare with the initial images). 10. Analyse V-slit images on "B" obtained in step 6 to get x~ij~ and I~ij~.

II. Slit "A" and screen "B" (or "C"). Expected beam parameters (350keV) at "A": x~rms~= ~x~\~ 2-4mm; ~x~ \~ 0.8-1.5 m; divergence (correlated) < 5mrad Expected beamlet parameters at "B" (from V-slit in "A") Beamlet divergence (rms): \~ 0.2-0.8mrad; x~rms~\~ 0.2-0.8mm (at L\~1m)

1. Set Q, U~DC~, R~0~, B~1~. Set: Buncher=OFF, B2=0 !!! (demagnetise !). 2. Steer the beam to the centre of "A" and "B" using steerers 01 and 06 only (not 02) 3. Insert V-slit "A". Check the image in "B" is vertical. 4. V-slit "A" = OUT. Measure Q with FC-01. 5. Store beam images on YAGs in "A" and "B" 5a. Store images of YAGs A and B with LEDs on (for image calibration purposes) 6. Scan V-slit "A" for a set of \{x~i~\}, i=10-15 and acquire slit images on YAG "B". 7. Measure Q with FC (to compare with the initial value) 8. Store beam images on YAGs in "A" and "B" (to compare with the initial images). 9. Analyse V-slit images on "B" obtained in step 6 to get x~ij~ and I~ij~.

III. Slit (V and/or H) "B" and screen in "C". Expected beam parameters (350keV) at "A": x~rms~= ~x~\~ 2mm; ~x~ \~ 2 m (buncher = ON) \[5\]; divergence (correlated) \~ 0 mrad \[5\] Expected beamlet parameters at "C" (from V-slit in "B") Beamlet divergence (rms): \~ 1 mrad; x~rms~\~ 0.2mm (at L \~ 0.2m)

1. Set Q, U~DC~, R~0~, B~1~, B~2~. Set: Buncher= ON or OFF 2. Steer the beam to the centre of "A" and "B" 3. Measure Q with FC 4. Store beam image from YAG in "B" 5. Scan V-slit "B" for a set of \{x~i~\}, i=10-15 and acquire slit images from YAG "C". 5a (optional). Scan V-slit "B" for the same set of \{x~i~\}as above and measure currents I~i~ with FC. (to compare I~i~ distribution with that obtained from scanning the CCD image) 6. Measure Q with FC (to compare with the initial value) 7. Store beam image on YAG in "B" (to compare with the initial images). 8. Store images of YAGs B and C with LEDs on (for image calibration purposes)

IV. Pepper pot in "B" and screen in "C" Since, the measurements with the pepper pot is a relatively quick procedure, measuring Q and acquiring YAG image after completion may not be required.

1. Set Q, U~DC~, R~0~, B~1~, B~2~. Set: Buncher= ON or OFF 2. Steer the beam to the centre of "A" and "B" 3. Measure Q with FC 4. Store beam images on YAGs "B" and "C". 5. Insert pepper pot and acquire image on YAG "C" 6. Store images of YAGs B and C with LEDs on (for image calibration purposes)

V. "Quick and dirty" procedure This simplified procedure is necessary for optimisation purposes during ERLP injector parameter scan. The transverse emittance is estimated using a pepperpot or slit in "B" and screen "C" according to :

where x and u are full widths of the beam (Screen "B") and the beamlet (screen "C") images; L is the distance between them and  is a coefficient translating the visible image width to the RMS width (=) as

.

With the saturated images  was found to be \~1.5-1.6 but this should be re-evaluated for the unsaturated images.

Photocathode lifetime Controls: laser pulses counter. Supplemented by Q measurements.

Off-axis beam generation Develop procedure for re-alignment and steering the beam with only those diagnostics that will be available in a final ERLP layout.

Injector optimisation procedure at given UDC, Q, and R0. Bunch length s :

is not particularly sensitive to BL parameters \[2\] apart from Q that may give variation of ~z~ within probably +/- 15%. DC gun voltage may have an appreciable effect but it is not feasible to deduce it from \[2\] because the B~1~ was kept constant during the U~DC~ scan that is not correct (over-focussing at lower U~DC~). First solenoid field scan at 250keV indicates however that the integral characteristics (emittances etc) are quite comparable (if not better) to those at 350keV. The latter is still much more preferable because of better phase-space distributions hence much better bunch characteristics at the output of the booster. 

B~1~ , R~0~ - does not have appreciable influence. B~2~ - ??? (no data but probably similar to effect of B~1~) P~b~– key parameter At buncher position, we should expect ~z~\~7-8mm (at ~0~=20ps). Electron energy spread E. B~1~, Q, U~DC~ – appreciably R~0~ – small influence B~2~ - ??? (no data but probably similar to effect of B~1~) P~b~ – no data available I assume the energy spread is of secondary importance compared to bunch transverse size, emittance and length. Hence it will not be considered in the optimisation process. Transverse emittance x. B~1~, B~2~ – key parameters P~b~ - as stated in \[Chris\], buncher has little influence on transverse phase space but I have not seen evidence yet (Boris to model later: 1^st^ cell position – with and without buncher) Q, R~0~ – strong influence Transverse size x. B~1~, B~2~ – key parameters R~0~ – may not have influence directly (with a proper choice of magnetic fields) but may conflict with achieving optimal emittance.

Procedure (extended – some steps may be omitted after a few first runs when general trends are established) 1. Set U~DC~, Q, R~0~. 2. B~2~=0, P~b~=0. Vary B~1~. Find minimal ~x~ on "A" and monitor ~x~ on "B" \[Plot ~x~(B~1~) for "A" and "B"\] 2a. Measure emittance at "A" at several B~1~ around an optimal value. \[Plot ~x~ (B~1~)\] 3. P~b~ =0. Vary B~2~. Get minimal ~x~at "B". 4. Measure ~x~ at "B" for several B~2~ around optimal. \[Plot ~x~ (B~2~)\] -> B~2~ optimised. 5. At B~1~, B~2~ = optimal, vary P~b~ . At each P~b~, ensure that the phase is set correctly (zero -crossing). Measure ~x~ and ~x~ on "B" -> plots. To establish an optimal P~b~ - use procedure given in "Bunch length measurements set-up". 5a. Compare ~x~ and ~x~ on "B" with and without buncher with computer simulations. 5b. Infer the bunch length ~s~ at "B". 6. Measure the energy spread ~E~.


Bunch charge measurement procedure- development (not applicable now) FC current measurements will be done with the voltage follower. The RF groups develops and implements it. FC could be also connected to the cable either directly or via a RC low pass filter.


The latter should have a cut-off frequency = 20MHz. The filter voltage gain:

If C>>C~F~, and the filter is not connected to the cable, . For f~0~=20MHz; Q=80pC (U~FC~=8V):

  1. R=110; C=70pF; ~charge~\~1ns; ~discharge~ \~ 3.5ns; U~c~=1.0V
  2. R=50; C=160pF; ~charge~\~0.5ns; ~discharge~ \~ 8ns; U~c~=0.5V


Since the FC does not have a proper FC geometry and the beam dump insert is made of tungsten, the secondary and backscattered electrons could present an appreciable error in charge measurements (potentially up to 50%). Introduction of some 50-100V biasing or use of the permanent magnets, at least on the evaluation stage, is desirable.

Also on Fandom

Random wikia