Nov 2009

Flux measurements at X25 and X29. These measurements were done with a photo-diode.
Tables

		We varied the apertures (slits) and measured the flux for a wide range of wavelength. Be aware for now, if you want to screen at longer lambda than 1.2A, it's fine but if you want to collect you should let us know in order to optimize the intensity and the shape of the beam. Normal mode is 50x50 um²--brown line and the flux is 2-3x10¹¹ photon/sec The dashed lines is to show the normalized to 100x100 um^{2 The native size of the beam (if we didn't have slits in) would be FWHM ~ 90 um vertical and ~260 um horizontal.}
		Comparison with X29 On this graph you can see the normal mode of operation for X25 (red) and the normal mode of operation for X29 (blue dash). At this point "normal mode" means different things for different beamline. slits size 50x50 um² vs 160x100um² divergence of the beam 0.3-0.5 vs 0.4 different beam characteristics 90-260 um² vs 130x260 um² The idea is that users can control most of these parameters and customize them to fit their need. ex. easy defocus procedure

26-27 August 2009 Horizontal Divergence Measurements

Scattering setup: The beamline energy was calibrated at the Se edge. The pass energy of the four bounce divergence analyzer (880) reflection was found at 13415 eV. The unit was set to scatter in the horizontal.

mono focus = 2.51 beam spot at sample (as measured with table scans) 344 um FWHM horz (dromedary) 85 um FWHM vert (dromedary) FEA 3000 um horizontal colli slits 20 um horz x 50 um vert E=13.407- 13.421 kev mirror bend cyl. =170,000steps		The following graphs show the divergence versus energy. Due to the strong sagittal focusing, a significant correlation between angle and energy is observed. Therefore, divergence scans were taking at energy intervals 2 eV apart through the energy width of the monochromator bandpass. The FWHM of the divergence envelope is ~317 urad. The right graph shows the integrated intensities at each energy, which gives the energy resolution; it is observed to be ~ 9 eV FWHM. These measurements are consistent
colli slits 100 um horz x 100 um vert E=13.407- 13.421 kev mirror bend cyl. =170,000steps		Same measurements done on June 2009 Divergence for focus at 2.51 ~380 urad, the horizontal slits in the hutch are different.

Then the sagital focus was brought to -0.4 in an attempt to get the second xtal flat. The data shows that there is only a very slight correlation between the horizontal divergence and the energy when the monochromator is defocused. The FWHM of the horizontal divergence is 42.3 urad.

Keeping the horizontal position the same, the effect of varying the vertical mirror focus was investigated. The vertical mirror bend was changed from 170,000 to 210,000 and the scans repeated. This value gave a somewhat better vertical focus. The results are identical. . A slight correlation between angle and energy is observed, and the divergence FWHM is 45.7 urad. To check this, the mirror focus was then changed back to 170,000, and an identical divergence scan was measured.

23-24 June 2009
Horizontal divergence of the beam measured with a four bounce crystal set in a dispersive geometry by Joe Dvorak.

The divergence measurements were made with a four bounce crystal arrangement set in a dispersive geometry. As such, the unit acts as an extremely narrow energy and angle filter. By rotating the entire unit about the center of the first crystal, the divergence of the beam can be measured directly. The rotational scan stage has a resolution of 0.33 urad/step, more than adequate for this beamline. The unit was set up to scan in the horizontal plane. The crystals are Si(220), incident angle on the crystals is 75^o to grazing, and the included angle between the dispersive crystals is 30^o.


Collimator slits = 100/100 um White beam vertical Front End aperture slit settings: 1.38 mm Photon energy = 10.057 eV near Zn		The following graph shows the dependence of the horizontal divergence on the white beam horizontal slit settings (FEA HORIZONTAL). The main graph shows the experimental data. The inset shows how the FWHM varies with the horizontal aperture. The results are reasonably linear over the range measured.
Front End Aperture Horizontal FEA=1.9mm		Next, we kept the white beam aperture fixed and varied the horizontal focus of the monochromator to measure the effect that the focus position had on the beam divergence. The measurements were done at several FEA HORIZONTAL settings: 1.9, 2.9, and 4.9 mm. The following graphs show the experimental data, with the insets showing the variation of the FWHM with the monochromator focus.
Front End Aperture FEA=2.9mm Normal Mode of Operation
FEA=4.9mm		The results show that when the beam is defocused (decreasing focus setting below the nominal “best” focus setting of 1.02 mm for this wavenlength), the divergence decreases approximately linearly with the focus setting, and can easily be decreased by a factor of two or more. This all comes at a cost in flux, of course. It was also observed that the flux (as measured by the “merit”) is directly proportional to the divergence. As the white beam horizontal aperture is increased, the FWHM of the decreases with defocusing as expected, however, the absolute value of the divergence maximum begins to vary with the focus setting. That is, the peak is always found within the expected envelope, but the exact position within that envelope becomes variable. It appears that the sagittal focusing crystal is very sensitive to any perturbations.

June18 2009.

After the May shutdown, X25 started with beam in the hutch despite huge work done on the mirror bender. Unfortunately the vertical beam looked doubled.
The paddle, right after the mono looked like it didn't have enough motions anymore to show the beam. It needed to go higher.
Shu, Lonny, opco , Kim Hayes, Marc and I (Annie) were there because we needed to unlock the padlock and have radiation people present while moving the paddle. Ultimately we didn't have to physically move the padlock and the beam was visible and looked like 3 orange stripes. We actually would like to see a flat fat beam. If it were to have an angle then we would have some issue with where the beam hits the first xtal. Not the case here.

We are at the selenium inflection point with slits of 100x100 um2 ( merit of 489 for I=217).
When we scan the vertical beam we can see two peaks. Moving the mirror in the vertical motion (initial position -0.88 to -0.68). after going up by steps of 0.2 , we see the two humps seperating and the intensity of both peaks is going down too.
Let's move down. eventually the two humps are about the same intensity.
We then turn the mirror bender (cylindrical) on and start moving it by 10 000 steps. The temperature of the motor (look on display on top of mirror tank) rises to about 44C when we move the bender. We end up unbending the mirror. We get a single peak with a lot more counts.
We move the wavelength from 0.9795A to 1.1A and we still have a nice single peak with, according to the scan, FWHM_v says 45um. Should be at best 50-60 um.
Final bender position: 170 000, height z=-1.78.

Jan 26 2009.

Tested with a diffraction pattern what the spot shape would look if we slit down the front end apertures.

B. Sweet--The beam-size tool I sent you on Friday predicted that the spot should be
160um fwhm with the FE slit at 4.9mm. Indeed that's what we saw for a
"typical" fully-recorded spot. However the spot got no smaller when we
narrowed the slit to 1.9 and 1.3mm, where the spot is predicted to be
60-75mu. What we >did< see, however, was that the peak:background ratio
appeared to increase from about 4 with the wide slit to 6 with the
narrower ones.

A. Heroux-These shots were not done taken with dose in mind to compare the intensity of the peak or the backgroung. As seen on the first row trippling the exposure time barely the backgroung but double the intensity of the peak.
On the second row the backgroup doubles but the paek at least triples (if not quadruples).


Huge xtal with slits at 20x20 um2 (not calibrated) 0.3 sec exposure time for front end aperture at 4.91mm(not calibrated).	Huge xtal with slits at 20x20 um2 (not calibrated) 0.6 sec exposure time for front end aperture at 4.91mm.(not calibrated).	Huge xtal with slits at 20x20 um2 (not calibrated) 0.9 sec exposure time for front end aperture at 4.91mm.(not calibrated).


Huge xtal with slits at 50x50 um2 (not calibrated) 0.3 sec exposure time for front end aperture at 4.91mm.(not calibrated).	Huge xtal with slits at 50x50 um2 (not calibrated) 0.6 sec exposure time for front end aperture at 4.91mm.(not calibrated).	Huge xtal with slits at 50x50 um2 (not calibrated) 0.9 sec exposure time for front end aperture at 4.91mm.(not calibrated).


Huge xtal with slits at 50x50 um2 (not calibrated) 0.3 sec exposure time for front end aperture at 4.91mm.(not calibrated).	Huge xtal with slits at 50x50 um2 (not calibrated) 1.0 sec exposure time for front end aperture at 1.91mm.(not calibrated).	Huge xtal with slits at 50x50 um2 (not calibrated) 3.0 sec exposure time for front end aperture at 1.31mm(not calibrated).

Jan 23 2008:
knife edge scans to determine the divergence of the beam at X25 while close the front end slits. Measuremnets done by Grace and Bob.
These results are with non-calibrated front-end slits and "20 um" apertures.
largest spacing is using lambda 0.9795A, a distance of 350mm for the detector apertures of 100x100 um2

Front-end slit width/mm	FWHM / mm	width – aperture	FWHM divergence / mrad	spot size / microns	largest spacing	relative intensity

1.31	0.0254	0.0054	0.0108	204	600	350
1.51	0.0413	0.0213	0.0426	215	500	650
1.71	0.0540	0.0340	0.0680	224	500	840
1.91	0.0699	0.0499	0.0997	235	500	1160
2.11	0.0794	0.0594	0.1188	242	500	1360
2.71	0.1080	0.0879	0.1759	262	400	1700
3.51	0.1365	0.1165	0.2331	282	400	1900
3.91	0.1588	0.1388	0.2775	297	400	2480
4.31	0.1746	0.1546	0.3093	308	400	2600
4.91	0.2381	0.2181	0.4363	353	300	3100

		λ / a = angular spot separation
		spot size = aperture + div.*dist.
		a = λ / (size/distance) * factor = λ / factor * [(aperture + div.*dist.)/dist.]

Summary of Tests performed on X25 during the week of 9/29-10/3/08:1 done by Grace)
Started in 2.584Gev Ops. The mono lift , mono height, mirror height, twists, roll and focus were realigned to optimized the beam shape on the mono and mirror phosphors.2) The mono lift and mono height adjustments successfully eliminated the double peak previously observed during pitch scans.3) The diffractometer was realigned to optimize counts. The diffyaw and diffpitch were optimized. The FrontEnd slits were closed up to decrease scattering.4) After the Energy was calibrated, the proper gap was measured. Chris then calculated the needed offset and set this in software.5) Preliminary settings for Roll and Focus were looked also looked at just prior to the ring energy switching back from 2.584Gev to 2.8Gev ops. 6) The mirror focus was also looked at. Although, the beam can be overfocused, attempts to underfocus the beam had unexpected results. Additionally the elliptical bend motion continues to give false hard limit trips. Long term our recommendation is to use the BC-005 (APS transition board) in place of Bill's IO58 box (in the white electronic rack), as it's wiring scheme has been more reliable. 7) The alignment was rechecked, the Energy was recalibrated, and the gap offset was optimized for 2.8Gev ops 8) The focus table was established for energies ranging from 6.0 to 15 kev, and the roll motion was further optimized.9) Bob and John L. made modifications to the periscope to accommodate a protective lens cover. This required the periscope motions and focal distances to be rechecked.10) A burn was taken on the Hutch Ion Chamber to confirm the beams position and the beam position monitor and attenuator were reinstalled. 11) On Friday Stu cooled down the detector and the cold stream was put back in place. 12) Joe D., Annie, and Chris worked on measuring the beam flux using Lonny's standard ion chamber. As the focus table was not fully implemented, hand tweaking was needed for these tests.13) Chris fitted the focus data to determine the best curve to calculate the focus for energy moves. Although this was not completely implemented until Friday evening, the focus table did appear to be working repeatably for .979 A and 1.1A. 14) As mentioned in my note at the beamline, the roll will still need some final tuning and the focus table will need additional testing. 15) If time allows, I would recommend completing the roll/focus testing with the existing slit calibration to maintain a consistent merit comparison..16) As the slits have had excessive use, the slit calibration will need to be rechecked prior to user operations. If new offsets are used please record these in the beamline logbook.

Summary of Sagital Focus values and expected Merits for changing Energy on X25:

								(10/6/08)

Energy	Wavelength	Theta	Focus	Roll	diffperp	optics_V	Colli	Hutch	Merit	Ring	H-FWHM	V-FWHM	diffpitch	diffyaw	slits

6.000	2.0664	11.40275	-0.300	-0.218	1.5215	-1.4300	128090	516824	738	170ma	243um		0.49	-0.36	100x100

7.113 (Fe)	1.7430	16.13710	0.020	-0.218	1.1217	-1.4400	92620	509328	541	169		*265	0.49	-0.36	100x100

9.000	0.6200		0.620	-0.280	2.8520	-1.3400	139000	485000	536	257	254	244	0.35	-0.37	100x100

10.000	1.2399	11.40275	1.020	-0.218	0.1419	-1.1200	118901	188708	651	~179	239		0.49	-0.36	100x100

11.000			1.420	-0.218	1.9789	-1.0470	139000	238000	865	161	231	152	0.33	-0.25	100x100

12.658 (Se)	0.9795	8.98580	2.150	-0.218	-0.3081	-1.0900	148501	141989	892	165	~236	*107	0.49	-0.36	100x100

13.777	0.9000	8.25126	2.650	-0.222	-0.3778	-1.1400	57386	65832	302	~186	258		0.49	-0.36	100x100

15.000	0.8266	7.57388	3.250	-0.218	-0.5783	-1.0600	75820	79801	407	184	239		0.49	-0.36	100x100

Bolded values reflect the most recent compatible "diffperp" positions using the same roll, diffpitch and diffyaw values.

Oct 3, 2008

After new upgrades of the motor controls for the mono.

There doesn't seem to have anymore flickering of the beam when we wlook at the phospho in the hutch. The motors ahve all been calibrated and we have a new algorithm for the focus.
Measurements of the flux by Joe Dvorak.
The photodiode is set at the sample position 10^6 V/amp, 4000 cv bias.
raw data

	C4 counts	Current uA from photodiode	flux (normalize at 235mA) photons/sec	ring current mA	Flux (to come)
0.9A 12658eV	100x100 0 200x200 0 300x300 0

0.9795 12658eV	100x100 18293 200x200 47887 300x300 71137	70.63 47.41 17.99	1.7x10^11 4.5x10^11 6.7x10^11	252.8--merit 726 252.8 249.8

1.0A 12658eV	100x100 18680 200x200 47263 300x300 68925	18.2 46.25 67.55	1.2x10^11 4.2x10^11 6.2x10^11	244.7--merit 802 244.7 244.7

1.1	100x100 27195 200x200 67918 300x300 102168	25.18 63.35 95.39	2.0x10^11 5.0x10^11 7.5x10^11	241.0---merit 1561

1.2	100x100 9604 200x200 24820 300x300 38174	8.11 21.10 32.35	5.6x10^10 1.5x10^11 2.2x10^11	238.1---merit 645

July 28, 2008
New support for the roll has been added. This is also a new second xtal in the mono.
We also have a new kappa arm called skinny kappa which we want to test. We believe we are at the Se edge and we are running in single lambda mode until motor issues are resolved.

DJ90_1 run 1 The beam seems fine on the phospho paddle. We have the slits at 100x100 just calibrated (with a graph?). This is 1.0 deg for 0.5 sec and it diffracts to 2.2A Rsym 0.078 0.1.66% rej. This data collection is mostly to evaluate the behaviour of the beam after the addition of a clamp on the roll. This xtal also has some Semet and I would like to see some signal in the data.
DJ90d4_4 Data collected at 0.9795A at X25 first graph no reject, secong graph after reject from the scan the infl should be at 0.9794A and peak at 0.9792A
at X29 0.4 sec for 1 deg, collected before mounting on X25. slits are at 100x 160 as usual. 0.9795A diffracts to 1.9A Rsym 4.6% rej 0.22%
This would be more usuful if it was at the peak and not the inf. The yellow and blue line should be apart for low res if there is some anomalous signal.

June 13, 2008 (it's a friday!)

After multi problems with the motors mostly of the mono, there is beam in the hutch.

Motors off (pitch , theta and lift) in other words we are at 1.0757A and we stay there.

I am collecting on a NAO mutant P3221 107 107 337 90 90 120.

Ring: digital feedback on (V and H)

ah24bb4 run 4 I can't really see the beam on the phospho paddle because it's to dim. When I open 200x200 then I can see a slight beam. the ring current is at 198.8mA and there was a global orbit correction right before I started collecting. This is 0.5 deg for 10 sec and it diffracts to 2.25A Rsym 0.061 0.12% rej the slits are 100x100. This data collection is mostly to evaluate the behaviour of the beam after the addition of a clamp on the roll during the shutdown. As you can see the scaling is very quiet except for a glicth at frame 100.

29-30 April 2008- with focus and defocused beam, slits 150x150

Ring: digital feedback on (V and H)

xtal_ke_5th (focus 2.41mm lambda 0.9795A) This is the same xtal as yesterday, not necessarily the same starting orientation. This is 1 deg for 1 sec and it diffracts to 1.5A as opposed to 1.9A for 0.5 deg 3 sec yesterday. Despite not a huge difference in the C2 and C4 counters, the intensity of the beam is obviously about 10x higher today. In the i0 plot, there are some spikes in the C4 (green) which are also seen at C2 (red). This scaling is what we used to have before the "clamp".
xtal_ke_5th (defocus 2.9mm lambda 0.9795A) A lot less intensity. Still some spikes in C4 and C2. It's not ovious if there is any improvement in the scaling caused by defocusing the beam.
xtal_ke_5th defocus with longer exposure time to pick more of the intensity exposure of 6 sec for 0.5 deg. The beam dumped at after 300 frames.
xtal_ke_5th 1.1A, focused 1.78mm. very weak intensity. We scanned the gap to make sure we were not too off. Tweaked the outboard twist to improve the profile of the beam. The C4 counter goes from 60K to 175K from 0.9597A to 1.1A but the C2 counter doesn't improve that much even when we refocused. The beam dies very fast and it is reflected in the scaling. According to Paul the ring gives beam extremely quiet today. Xtal barely diffracts to 2.1A Rsym=6.5% reject 5.5% for 360 frames. We have to keep in mind that we didn't have time to really optimize X25 (move the mono horizontally??).

28 April 2008- with focus at 2.4 and slits 150x150 (because I am missing the xtal)

xtal_ke_5th This is for 0.5 deg 3 sec exposure no attenuator Rsym 6.0 rejects 0.19% The temporary clamp on the xtal-roll has been removed killed the previous xtal, so I had to take a small one not collected on x29. this scaling IS as quiet as X29 ?????? The green line on the I0 plot shows a bigger drop than expected. We did not move the roll motor after being unclamped. The pitch motor seems to have some reliability problems.
xtal Ke 10ai-1 collected on april 18 2008 on X25 same xtal as below 0.5 sec with one attenuator xtal shot at the tip	no I0 plot because the data collection stopped (too short exposure time)
xtal Ke 10ai-1 (Ke 10ai-1_3 collected on april 2008) on X29 for this data collection, the scaling is pretty much typical for this beamline.

Here are some of the datasets collected -1 march 2008- with focus at 1.8 and slits 100x100

This is for 0.5 deg 1.25 sec Rsym 5.6 rejects 2.2%
This is for 1 deg 2.5 sec Rsym 5.3 rejects 2.5%
This is for 2 deg 5 sec Rsym 6.3 rejects 2.18%
So I adjusted the exposure time to be about equivalent per degree. That seems to be the major factor as far as rejection. What really bothers me is the noise in my scaling. Here are some scaling for the SAME xtal at X29 (slits v=100 H=160) The scalings look the same and are very smooth. They always are.	X29 1deg 3 sec Rsym 6.3 reject 3.8%	X29 2 deg 1 sec Rsym6.4 reject 1.5

Data collected at X25 on Feb 28, 2008 defocused at 1.8 slits 100x100

This is for 0.5 deg 10 sec Rsym 7.1 rejects 15%
This is for 1 deg 10 sec Rsym 7.7 rejects 8%
This is for 2 deg 10 sec Rsym 7.5 rejects 2.8%
So I adjusted the exposure time to be about equivalent per degree. That seems to be the major factor as far as rejection. What really bothers me is the noise in my scaling. Here are some scaling for the SAME xtal at X29 1 deg for 3 sec Rsym 6.3 reject 3.8% 2 deg for 1 sec Rsym6.4 reject 1.5 the scalings look the same and are very smooth. They always are.

Data collected at X25 feb 28, 2008 focussed at 2.8 slits 100x100

1 deg 2 sec Rsym 6.0 Rejects 2.2% (6Hb_x25_3)
2 deg 2 sec Rsym 7.0 Rejects 0.9% (6Hb_x25_6)
2 deg 2 sec Rsym 7.0 Rejects 0.9%(6Hb_x25_7)

May 2008 shutdown: clamp was removed and, to substitute for its impact on suppressing vibrations while allowing freedom of motion of the second crystal assembly roll cradle, the cradle was re-built, i.e. it was stiffened, springs were replaced, and flex pivots were replaced. This succeeded in substituting properly for the removed clamp, in terms of eliminating vibrations, while allowing for freedom of motion of the cradle.

Winter 2008 cycle: the early part of the cycle was given to commissioning the new micro-diffractometer, which had its own set of issues that needed to be dealt with. Once they were dealt with, continued studies of the vibrations were undertaken. A horizontal beam instability, in the ring, was discovered, affecting not only X25 but also X29 and other beamlines, due to the digital feedback system not being tuned up. As it happened, this instability, although evident in frequency analysis of the x-ray beam (at around 20 Hz), was too fast to cause the ~2 Hz effect that we knew was in the beam and likely associated with the monochromator. At Lonny Berman’s suggestion, the second crystal assembly roll cradle was clamped, and that appeared to eliminate this ~2 Hz nuisance entirely, although the 20 Hz ring instability in the horizontal direction was still present. Test crystallography data confirmed that the ~2 Hz nuisance was the likely cause of all the earlier problems for crystallography data collection, because those problems entirely disappeared upon installation of the clamp.

December 2007 shutdown: vibration isolation pads were added, beneath the monochromator support, per Ed Haas … no impact at all on the vibrations, when beam resumed operation in January 2008.
May 2007 shutdown: open up monochromator chamber, and immediately “a finger was put on the problem”, literally speaking. The second crystal assembly roll cradle, in particular, was felt (with a finger) to be vibrating. This got confirmed using an accelerometer. A dampening weight was added to this portion of the assembly. When beam resumed operation, this showed marginal improvement in the vibrations.

More tests and measures done during early summer 2007: shut off all potential vibration-inducing systems at the beamline such as pumps, the cryogenic cooling system, even the building air conditioning systems in the nearby mechanical equipment room … no impact at all on the vibrations. Also installed added supports beneath the monochromator (per Mary Carlucci-Dayton’s advice) and isolate it from the upstream portion of the beamline (by undoing a couple of clamps surrounding a bellows) … no impact at all on the vibrations. Added another dampening weight to the second crystal assembly roll cradle and changed its retaining springs, which again led to marginal improvements in the vibrations, but still they were observable. Nevertheless, they were at a small enough level that reasonable crystallography data could be collected, and thereafter users were booked to collect data at X25 from July through the end of the 2007 calendar year.

Spring 2007: new optics installed and commissioned, vibrations (beam flickering) noticed right away, particularly in the horizontal direction.