Argonne National Laboratory

Advanced Photon Source
Sector 7

Argonne Home > Advanced Photon Source >

MHATT-CAT Monochromator Upgrades during the October shutdown


Eric Dufresne, Dohn Arms
MHATT-CAT
November, 2001
(www.mhatt.aps.anl.gov/Sectors/Sector7/Operations/FY02_run1/)



Table of Content:

Introduction.

The work shown below is being done to improve the stability of our LN2 cooled 
HHL Monochromator. We have found when running in non top-up mode that the
monochromatic beam position moves substantially over a fill, and the motion
is periodic from fill to fill. Some heating effects cause beam motion, and it
will be important to stabilize this motion in non top-up mode. The beam is
quite stable in top-up mode, but the some heating period is required before the
beam returns to its original position.  This heating effect is present in all
operation modes, with the beam being most stable in top-up operations.

During the last shutdown, we added a braid to cool the second monochromator 
crystal. This braid will help to match the two crytal's lattice parameter which
thus should correct an energy dependent beam motion far away from the 
monochromator. We hope it will reduce heating effects on the second crystal too.
Below are the results from studies done in November. But before studies, a
summary of the work done with pictures is shown in the next section.

Summary of changes made to the 7ID monochromator during the October 2001 shutdown.

During the last few days of the October shutdown, Dohn Arms and Eric Dufresne
realigned the first crystal in the monochromator tank. The procedure goes as 
follow. The horizontal slide of the second crystal was first levelled
(see Fig. 10-31.1).  Then the first crystal tilt was levelled
(see Fig. 10-31.2. The chi offset unfortunately changed
to 0.4 degrees from 0.1 degrees, and no adjustment exists on the chi angle
except with the seal (see Fig. 10-31.3).  The current
chi angle will steer the beam a little  outboard.  Then the offset between the
two crystals was readjusted to make the vertical offset from the white beam
35 mm. A plastic jig of known thickness is inserted between the two crystal when
their Bragg angle is set to 15 degrees (Fig. 10-31.4).

To reduce the second crystal thermal drifts, we installed also a new braid. 
(see Fig. 10-31.6). We initially also added a Compton
shield, but decided not to keep it inside for this run. After all these changes,
the tank was pumped, but a leak was detected on the first crystal seal. The
chamber was reopened, and with a leak detector connected to the LN2 lines, the
leak was sealed. The base pressure of the first crystal tank is now 
7x10^-9 Torr, nearly three times as low as during the last run.
Realigning the H-slide.

Fig. 10-31.1. Final adjustment of the horizontal slide of the 7ID monochromator.

Realigning theta.

Fig. 10-31.2. Final adjustment of the first crystal theta tilt angle of the monochromator.

Final chi tilt.

Fig. 10-31.3. Final adjustment of the first crystal chi tilt angle of the monochromator.

Realigning the offset.

Fig. 10-31.4. Final adjustment of the two crystal offset.

The monochromator reopened.

Fig. 10-31.5. Eric, and blue gloves.

The monochromator reopened. Note the new braid.

Fig. 10-31.6. The monochromator after changes made in the October shutdown.
Note the two new Cu braid cooling the second crystal.

Saturday, Nov. 3, 2001. Time series with a 2 mm (H) by 1 mm (V)
white beam at 6.9 keV with a braid.

Fig. 11-3.1 shows the fluxes in 7ID-B and C during a
16 hour long time serie. The flux in 7ID-B recorded by an ion chamber (IC B) 
shows a sudden chnage 30 minutes in the time serie due to a tweak of the second 
crystal. Ignore the flux in the 7ID-C ion chamber since it was not in the beam 
(re: IC C). The flux in the 7ID-C hutch is detected by the BPM diode array at 
the entrance of 7ID-C 49 m from the source.  The diode sum (diode) shows 30 %
variations over the fill. The 7ID-B sees +10/-15 % changes not quite 
proportional to those seen in 7ID-C but clearly correlated with the fill
pattern (IRing) occuring every 12 hours. The LN2 fill occur every two hours or
so.
Intensity data on 11/03/01.

Fig 11-3.1 . Time series of the beam intensity, in 7ID-B and 7ID-C, starting at 1h00 am on 11/3
and lasting 16 hours. Also shown, the ring current decay, and LN2 level.

Fig. 11-3.2 shows the beam position in 7ID-C, 49 from 
the source.  After the tweak, the beam moves significantly by 0.4 mm and 1 mm in
the horizontal and vertical direction. The braid did not help significantly 
since this type of motion was also observed last September.

Fig. 11-3.3 shows the temperature of the second
monochromator crystal during the same time series. There is a significant 
warming effect of 12 C during the fill. With a 3 mm by 2 mm white beam on the
first crystal, the second crystal heated up by more than 30 C. Without beam, 
a thermocouple secured on the second crystal mount reaches -100 C or so. 
Clearly more work needs to be done to understand where the heating effects 
occur.
XBPM data on 11/03/01.

Fig 11-3.2 . The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C during the same time serie as Fig. 11-3.1.
The time serie is started on 11/03 at 1h00 am, and lasts for 16 hours or so.
After 30 minutes, I retweaked the second crystal and the beam moved vertically
by 1.2 mm. Large vertical motions (1 mm) occur during the fill as the beam current
decay. The beam is stable to 0.4 mm horizontally.

XBPM data on 11/03/01.

Fig 11-3.3 . Time series of the second crystal mount temperature during the fill. It changes by 12
Celsius during a fill with a 2 mm (H) by 1 mm (V) white beam on the mono set to 6.9 keV.
The initial decay is due to the fact that the white beam slits were set to 3 mm
by 2 mm and then closed before starting
the time series.

Friday, Nov. 16, 2001. Time series with a 0.5 mm (H) by 0.5 mm (V)
white beam, with the mono at 12.5 keV with a braid.

During the alignment period of Daniel Core's experiment, I ran time series of
the monochromatoc beam position and intensities at 12.5 keV, with tight L5-20
slits settings of 0.5 mm by 0.5 mm. Fig11-16.1 shows
the intensity versus time. After top up resumed, a periodic oscillation of the 
7ID-B and C ion chambers as well as the diode signal is observed with a 1.3 
hours period. Some small beam motion in Fig11-16.2
correlates with these intensity variations. The pressure changes in the Oxford
cooling system move the beam vertically by 10 microns. The beam is quite stable
although some fast noise is clearly present. For this measurement, the diode is
slightly off-centered by about 1 mm.
Intensity data on 11/16/01.

Fig 11-16.1 . Time series of the beam intensity, in 7ID-B and 7ID-C, starting at 1h17 am on 11/16
and lasting 8.5 hours. Also shown, the 7ID-C diode sum, the ring current decay, and LN2 level.

XBPM data on 11/16/01.

Fig 11-16.2 . The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C during the same time serie as Fig. 11-16.1.
The time serie is started on 11/16 at 1h17 am, and lasts for 8.5 hours.
After 1.25 hour, top-up operations resumed, resulting in a sudden motion.
The beam is stable about 10 microns in top-up operations.

Saturday, Nov. 17, 2001. Time series with a 0.5 mm (H) by 0.5 mm (V)
white beam, with the mono at 12.5 keV with a braid.

Fig11-17.1 shows the intensity versus time with the 
same conditions as yesterday. The 1.4 hours periodic oscillation is still 
present in all the ion chambers, either in the 7ID-B or C hutches or the 7ID-C
BPM diode. Note that the 7ID-C ion chamber sudden drop of signal near t = 7.25
hours is due to the experimenter closing slits downstream of the BPM diode.

Fig11-17.2 shows the BPM data. The long term stability
during this time serie is about 5 microns. Note the very slow drift from 
yesterday's data in Fig11-16.2. The beam has drifted
15 and 10 microns in the hor. and vert. direction respectively. Note also 
that the BPM was offset by 0.55 mm (V) and -1mm (H) in the plots, so we are 
at most 1 mm offcentered on the X-BPM and the calibration is for a different
energy either 7 or 10 keV. The calibrations can vary by a few percent, 
depending on the beam size on the BPM and the X-ray energy.
Intensity data on 11/17/01.

Fig 11-17.1 . Time series of the beam intensity, in 7ID-B and 7ID-C, starting at 2h17 am on 11/17
and lasting 8.25 hours. Also shown, the 7ID-C diode sum, the ring current decay, and LN2 level.

XBPM data on 11/17/01.

Fig 11-17.2 . The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C during the same time serie as Fig. 11-17.1.
The time serie is started on 11/17 at 2h17 am, and lasts for 8.25 hours.
The beam is stable about 5 microns in this shift of top-up operations!

Saturday, Nov. 18, 2001. Time series with a 0.5 mm (H) by 0.5 mm (V)
white beam, with the mono at 12.5 keV with a braid.

Fig. 11-18.1 shows a time series of the intensities 
in 7ID-B and C. The behavior is similar to data from the last few days. Still 
present is the intensity oscillation with a 1.4 hours period.

Fig. 11-18.2 shows the beam positions for the same 
time series. The long drift that was seen in the last few days has disappeared,
and now the beam is very stable on an half a day time scale i.e. the DC values
are the same as yesterday at the time of the time series. The largest beam
motion (10 microns) occurs when the cryocooler fills.
Intensity data on 11/18/01.

Fig 11-18.1 . Time series of the beam intensity, in 7ID-B and 7ID-C, starting at 0h13 am on 11/18
and lasting 10.5 hours. Also shown, the 7ID-C diode sum, the ring current decay, and LN2 level.

XBPM data on 11/18/01.

Fig 11-18.2 . The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C during the same time serie as Fig. 11-18.1.
The time serie is started on 11/18 at 0h13 am, and lasts for 10.5 hours.
The beam is stable about 10 microns in this shift of top-up operations!

Tuesday, Dec. 11, 2001. Time series with a 0.5 mm (H) by 0.5 mm (V)
white beam, with the mono at 10.0 keV with a braid.

In preparation for regular operation in the non-top-up Hybrid mode, I took some 
data with the X-ray Beam Position Monitor in 7ID-C. One expects the beam to be 
less stable in the hybrid mode when compared to the low-emmittance top-up mode.
The monochromatic beam energy was set to 10.000 keV and the white beam slits
were cut to 0.5 mm by 0.5 mm. 

Fig. 12-11.1 shows a time serie lasting about one half
hour, with sampling every seconds. The beam was stable to 1 micron RMS in both
the X and Y direction. The average position in X was 0.361 mm and in Y -0.015 mm
. 
XBPM data on 12/11/01.

Fig 12-11.1 The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C. The time serie is started on 12/11 at 0h34 am
and lasts for about 0.5 hours. The beam is stable to a few microns in
this shift of top-up operations!

Thursday, Dec. 13, 2001. Time series with a 0.5 mm (H) by 0.5 mm (V)
white beam, with the mono at 10.0 keV with a braid in the new Hybrid Mode.

Following the change over to the new mode of Hybrid operations, Don Walko 
suggested to measure time series of the beam position in 7ID-C. The results 
below in Fig. 12-13.1 shows that the diode sum, 
normalized to the beam current decay, is stable to about 10 % over 18 hours. 
The LN2 fill occurs every three hours and can be correlated with beam motion in
Fig. 12-13.2. The beam vertical and horizontal position
is stable to about 20 (V) and 30 um (H). An horizontal bump occurs when the 8 am
fill on fill occurs. Yizhak is not worried about the horizontal motion, and is
not too concerned with a 20 um vertical motion. One hundred microns would be
more troublesome for their experiment. 20 um, 19 m from the monochromator 
corresponds to a  1 microradian tilt.
Intensity data on 12/12/01.

Fig 12-13.1 . Time series of the beam intensity in 7ID-C, starting at 17h20 on 12/12
and lasting 17.5 hours. The 7ID-C diode sum, the ring current decay, and LN2 level
are the only valid data because both ion chamber in B and C were disabled.

XBPM data on 12/12/01.

Fig 12-13.2 . The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C during the same time series as Fig. 12-13.1.
The time series is started on 12/12 at 17h20, and lasts for 17.5 hours.
The beam is stable to about 20 and 30 microns in the vertical and horizontal
direction respectively during this shift of hybrid operations. Note that the large dips
are due to hutch access in 7ID-C.

Monday, Dec. 17, 2001. Time series with a 0.5 mm (H) by 0.5 mm (V)
white beam, with the mono at 10.0 keV with a braid in the new Hybrid Mode.

The data was started on 12/13 at 13h22 and lasts for more than 4 days. Three
figures are shown below. Fig. 12-17.1 shows a time 
series of the intensity not normalized to beam current decay. the data was taken
every ten seconds.  The diode signal tracks the ring current decay as expected.
Gaps occur in the data because of access to 7ID-C. The LN2 fills occur every 3
hours or so. The 7ID-B and C ion chamber were disabled for this tseries. 

Fig. 12-17.2 shows a time serie of the 7ID-C Beam 
Position. The beam motion peak to peak is 42 and 60 microns in the vertical and
horizontal direction respectively. This was acceptable to Yizhak's experiment
but would cause problems in an XPCS experimental with very slow time constants.

Fig. 12-17.3 shows a time serie of the second crystal
temperature showing a 1.5 Celsius change of the second crystal temperature 
during an APS fill. Overall, the hybrid fill and non top-up operation modes in
general cause much larger beam motion but for small white beam incident on the 
mono, the motion can be made small enough for many experiments at MHATT-CAT.
Intensity data started on 12/13/01.

Fig 12-17.1 . Time series of the beam intensity in 7ID-C, starting at 13h22 on 12/13
and lasting 4 days. The 7ID-C diode sum, the ring current decay, and LN2 level
are the only valid data because both ion chamber in B and C were disabled.

XBPM data started on 12/13/01.

Fig 12-17.2 . The beam position, 49 m from the source, or 19 m from the
High Heat Load mono in 7ID-C during the same time series as Fig. 12-17.1.
The time series is started on 12/13 at 13h22, and lasts for 4 days.
The beam is stable to about 42 and 60 microns in the vertical and horizontal
direction respectively during this shift of hybrid operations.

Temperature data started on 12/13/01.

Fig 12-17.3 . Time series of the second crystal mount temperature during the
4 day long tserie. It changes by 1.5 Celsius peak to peak during an APS fill.

U.S. Department of Energy UChicago Argonne LLC Office of Science - Department of EnergyOffice of Basic Energy Sciences - Department of Energy
Privacy & Security Notice | Contact Us | Site Map