MHATT-CAT Useful Operational Data for Run 2 of FY04

written by MHATT-CAT staff.
MHATT-CAT
Started Jan. 19, 2004
(www.mhatt.aps.anl.gov/Sectors/Sector7/Operations/FY04/run2/)

• Introduction
• Summary of the shutdown activities.
• Thursday, Jan 29, 2004: First day of beam for run 2 of FY04.
• Friday, Jan. 30, 2004: Cu shields shimming.
• Saturday, Jan. 31, 2004: white beam slit motor problems.
• Sunday, Feb. 1, 2004: steering correction test.
• Monday, Feb. 2, 2004. Summary of start up activities by
• Wednesday, Feb. 4, 2004. White beam filter by thin Si wedge
• Wednesday, Feb. 11, 2004. Recalibration of the monochromator energy by Don Walko.
• Tuesday, Mar. 2, 2004. New temperature screen from Dohn Arms
• Wednesday, Mar. 10, 2004. Verification of the horizontal white beam alignment by the survey group
• Thursday, Mar. 11, 2004. Several time series and fix the userCalc10.
• Friday, Mar. 12, 2004. Several time series and effect of the cryocooler pressure set point on beam position!
• Saturday, Mar. 13, 2004. Several time series.
• Sunday, Mar. 14, 2004. Time series of the atmospheric pressure and first crystal temperature.
  • Effect of atmospheric variation on T1
  • Beamline time serie.
  • Repair of the YBPM axis.
• Tuesday, Mar. 16, 2004. Main beamline IOC network problems.
• Wednesday, Mar. 17, 2004. Hutch 7ID-C IOC problem.
• Sunday, Mar. 21, 2004. Time series analysis.
• Wednesday, Mar. 24, 2004. New users, slight realignment.
• Sunday, Mar. 28, 2004. Cryocooler buffer level low.
• Tuesday, Mar. 30, 2004. Cryocooler buffer fill.
• Thursday, Apr. 1, 2004. Time series at 15 keV.
• Monday, Apr. 5, 2004. Solenoid replacement.
• Tuesday, Apr. 6, 2004.
• Friday, Apr. 9, 2004. Time serie, and bunch pattern single bunch mode.
• Monday, Apr. 12, 2004. Time serie, at the end of the single bunch mode.
• Lessons learned, to do list.

Introduction.

The work shown below is an account of useful operational data taken during
run 2 of FY04 on 7ID. This page will contain useful stability information on the
7ID High Heat Load Monochromator and also on the various repairs and problems 
identified during the run.

Summary of the Jan 2004 shutdown activities.

The APS upgraded several pressure sensors on their PSS DI-Water circuits
during the first week of January.  This affected all sectors.
See these pictures for details.


The 7ID-A crocooler was serviced this shutdown. The bearings were replaced,
a leak on the high pressure side was found on the high pressure buffer level
sensor and repaired, and the low pressure fill modification was removed.
Preliminary result show that the beam may be more stable with this modification.


The 7ID-C Kappa was tested for damage by technicians from MicroControle.
No serious damage was noted to the circles ball bearings.  They disassembled
the unit and packaged the two bottom circles with the base to be shipped back
to France.  See the full story here.



Dohn Arms upgraded the motor crate to the latest stable release of EPICS.
The crate now is equipped with load and go so that scans can be set up more 
quickly. Dohn Arms also upgraded the beamline control IOC to the latest 
stable EPICS.  It required a tremendous effort from Dohn to get this done!

Toward the end of the shutdown, the APS survey team started to check all
main beamline components for position. Don Walko brought them to 7ID to
investigate the source of the beam blockage that we have seen in the downstream
hutches after the Micromonochromator in 7ID-B. They found that the L5-20 height
was unchanged, corrected a small table misalignment of 0.5 mm horizontally,
found the commissioning window was centered, the L5-20 were centered reasonably
well.  In 7ID-B, the micromonochromator table was set also well. On Thursday,
Jan 29, Don Walko and Eric Landahl opened the Micromono to see whether the beam
blockage on the Monochromatic height at 1435 mm is caused by some object located
inside the Micromonochromator. They found that the Cu masks on the input and
output slits used as Compton shields were very close to the monochromatic beam
and shimmed these parts to raise them by 3mm.  See the full story here.

First day of beam time, Thursday Jan 29, 2004.

From D. Walko, edited by ED
The beam was delivered by APS as planned on Thursday 8 am. The beamline shutter
was off since Don Walko brought the APS survey and alignment team to investigate
the blockage of the monochromatic beam (from the high-heatload mono, not the
micromono) in pass-through mode. In the past we have observed that the top of
the HHL mono beam is sometimes clipped by an edge that we thought was in the
micromono.  

On Thursday the APS survey team checked the alignment of the white beam slits
and front end mask in 7ID-A. The commissioning window table was at the correct
height but the upstream part of the table was moved by 0.5 mm inboard. The L5-20
chamber was at the nominal position provided that they used the alignment number
written on the chamber by Oxford. They also checked the micromonchromator
alignment in 7ID-B and it looked close to its nominal position.

Today we opened up the micromono chamber and found the very likely source(s) of
the edge(s): some copper housing around the slits.  More details, including some
pictures, are posted on my website , so please look there for clarification.
It is not clear at this point what the purpose of these copper housings is, or
what can be done to move them. Can we simply shim them up a bit, or would that
affect their function?  I was unable to find these housings in the mechanical 
drawings of the micromono which were available to me, so I'm somewhat puzzled
as to their purpose.

The APS survey crew will return tomorrow morning (Friday) to confirm whether the
center of rotation of the crystals is at the white-beam height, and they should
also be able to tell us whether the things around the slits are really at a
height which could affect the HHLM mono beam. 

Friday Jan 30, 2004:shimming the Compton shields of the micromono slits.

From Don Walko, edited by ED:
The APS surveyors determined that in the micromono, the bottom of the copper
structures around the slits were at a height of 36.25 mm above the white beam.
While the nominal height of the mono beam is 35 mm above the white beam, things
such as steering errors, detuning, and simply the height of the beam make this 
tolerance too small.  Therefore we are reasonably confident that these blocks
were clipping the top of the mono beam in pass-thru mode.  Jon Tischler stopped
by from UNI-CAT and figured they were Compton shields and that it would be fine
to shim them up.  I placed an emergency order to the APS machine shop for little
blocks made from OFHC copper that was lying around and installed them this
afternoon.  Eventually we may want to replace them with better-sized pieces of
Cu.

The micromono has been slowly but steadily pumping down; the pressure is now in
the mid-7's.  Its base pressure--before the new, small Cu blocks were
installed--was in the mid-9's.  Since we needed a "type C" radiation survey
for the minihutch, I had that done and got the beamline up and running.  I got
the mono peaked up within one minute, which means it hadn't moved much after
being warmed up for cryocooler maintenance and having had its vacuum opened.  I
am leaving beam on in the A hutch to keep the crystal warm.

Tomorrow, I will image the beam in the B hutch to see if I can still find any
edges (hopefully not).  Assuming that the micromono pressure has continued to
fall, then when nothing else is going on, we will put beam into the C hutch and
use the beam position monitor for stability studies.  Our early indications are
that undoing the modification to the cryocooler will help it be more stable, but
of course we need to observe the beam to be sure.

Overall, I'm optimistic that the stability of, and our understanding of, the
beamline is improving.

PS by ED
It looks like we found one of the source of clipping on 7ID.  Initial
experiments on the beamline found as early as 1998 that we were short in flux by
about a factor 2.  In March 2001, we found how to clear this aperture with a
steering correction and gained a huge factor in intensity.  Today Don and his
team found out that the top plates in Cu on top of the slit assemblies (see web
site below) are very close to the mono beam, within 1.25 mm actually (36.25mm
from white beam height). Since setting the mono offset is not an exact thing (I
easily can misalign it by .5mm) and the fact that due to the temperature 
difference between the first and second mono crystal, the beams moves up, these
Cu mask were designed to be to close to the beam. (The second crystal being
warmer has a larger d-spacing thus at fixed energy a smaller Bragg angle. It
thus does not return the beam to the horizontal direction and the beam moves
up.) See Don's page for pictures.

Note that the beamline was down for 5 shifts at the beginning of the run until
Friday night due to the presence of the Survey team from APS. They needed access
to 7ID-A on Thursday, and to 7ID-B on Friday. The micromonochromator was opened
on Thursday afternoon.

Saturday, Jan. 31, 2004:L5-20 problems.

Just as we were getting everything figured out with the MHATT-CAT
micromonochromator in 7ID-B, we have encountered a problem with the white-beam
slits ("L5-20") in 7ID-A. To bring everyone up to speed, the horizontal blades
are not moving well if at all.  In fact it's the (five-phase) motors which don't
move well even when the shaft is disconnected from the slit.  Eric Landahl and
myself are fairly certain that it's the (fairly obscure) motor driver that is
wearing out, that the company (Berger Lahr) doesn't even advertise on their
website.

The current status is that I got beam thru the slits and am about to start a
time-series to check on the stability of the HHL mono.  The cables to the slit's
horizontal are disconnected so no one can accidentally (try to) move the blades.
On Monday we will either have to find a replacement driver or else begin wiring
up the motors to five-phase ACS Step-Pak drivers (we have three on hand).  This
may delay the start of the next beamtime, depending on how severe the problems
are (and assuming that the motors themselves are fine).  
D. Walko, edited by ED.

On Jan 31, 2004 at 7pm, Don Walko started a time serie with the white beam slits
set to ~0.3 mm (V), and uncalibrated in the horizontal direction. The undulator
gap was set at 21.755 mm(10.05 keV), and the flux was 3.7 x 10^12 ph/s in 7ID-B.
The beam was on the monochromator for well over 12 hours before the time serie
started. As seen in Fig01-31.1, the APS is running in
non-top-up mode, with the 324 bunches, and the fill-on-fill occur around 
t=0.75h. The 7ID-B ion chamber tracks well the ring current, but the 7ID-C diode
sum increases monotonically over the period, likely because the He flight paths
are become purer, oxygen and nitrogen free from the He flushing.

The beam position shown in Fig01-31.2 after the fill
seems to stabilize. Note the horizontal and vertical motion near t = 3.7 h
caused by the cryocooler fill. Note also the horizontal bump occuring after the
fill, likely caused by a slight orbit change. Finally, note that the vertical 
beam position has a very fast and large amplitude. This is new and we've never
seen this before. This will need to be looked into in future studies time.

We also took some oscilloscope scans of the beam position monitor signals and
they have not yet been analyzed.

Fig 01-31.1. Time series of the beam intensity in 7ID-C, starting at 19h on 01/31/4 and
lasting 4 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is not connected.

Fig 01-31.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 01-31.1.

Sunday, Feb. 1, 2004: steering correction tests.

After tweaking the HHL momo around 11pm on 1/31, a time serie was started with
the monochromator set to 10 keV, and 7ID set to 21.755 mm (10.05 keV). 

Fig02-01.1 shows the intensities in the non top-up 
mode. Note that the 7ID-B ion chambers follows the ring current. Due likely to
He purging out the air in the flight paths, the 7ID-C diode signal increases
even though the ring current decreases. The diode sum signal finally follows
the ring current after t = 6h. So it takes quite a bit of time to purge the He
flight paths. Note the funny noise slightly before the fill-on-fill near t = 7h.

Fig 02-01.1. Time series of the beam intensity in 7ID-C, starting at 00h30 on 02/01/04 and
lasting 9 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is not connected.

Fig02-01.2 shows the beam positions in the same time
series as the previous one. The horizontal beam motion due to the filling of the
low pressure LN2 vessel in the cryocooler causes beam motion around 40 microns
peak to peak. Note the horizontal beam motion during the fill-on-fill of 20 um.
The vertical motions seems to be much larger than before with an rms deviation
on the order of 10 microns. The last two hours of data has an RMS deviation of
13 microns. It is possible that either the beam is moving more than it use to
or the conditioning electronics has changed. We have upgraded the 7ID-A crate
so it is possble that the AD board settings have changed. This needs to be 
investigated further.

Fig 02-01.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 02-01.1.

Following this time series, Eric Landahl performed a steering correction test to
see whether an edge is still present in the beam after the micromonochromator
modification. Before the test, the L5-20 top blade was moved up by 1.5 mm so 
that the beam after a nominal 25 microradian orbit correction would pass through
the slit opening. 

Fig. 02-01.3 shows the steering angle in a time series, the angle moving up by
25 microradians up, then down.Fig. 02-01.4 shows the flux measured in 7ID-B 
and 7ID-C during the steering two corrections as a funcrion of time. The first
rise in intensity after 0.2 h is due to the fact that the integrated intensity 
passing through the L5-20 has increased following the steering correction since
now most of the beam clears the bottom blade of the L5-20. This sharp rise is 
followed by a drift which is likely due to beam current decay and monochromator
heating. The sudden drop of intensity near 0.9 h is not understood. The sharp
drop in intensity at 1.25 h is the second orbit correction. 

a href="#Fig02-01.5">Fig. 02-01.5 shows the beam position during the orbit
correction. The beam moves up in 7ID-C by 1.2 mm, or 1.2 mm/49.2 m = 24.4 
microradians. The APS BPM data are consistent with the 7ID-C BPM. Note that
the horizontal position shifts by 0.1 mm. This is possibly caused in part by 
a misalignment of the 7ID-C BPM axis. Also the orbit correction was done in two
steps, the second step is not present on the APS BPM data. This is the first 
measurements we've done that shows the correlation between APS and 7ID-C BPM 
data. Note that the rapid dips near t = 0.6 and 0.8 hrs are hutch access to take
burns.

Fig 02-01.3. The steering angle versus time.

Fig 02-01.4. Time series of the beam intensity in 7ID-C, starting at 09h28 on 02/01/04 and
lasting 1.6 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is not connected.

Fig 02-01.5. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 02-01.4.

After this study, Eric Landahl changed the monochromator energy to 18 keV. 
Several hours afterwards, at 16h36, after the monochromator was retuned to 10
keV, a time series was started. Fig. 2-01.6 shows the fluxes while Fig. 2-01.7
shows the beam position. Probably due to the change in monochromator energy, 
the flux peaks 1 h into the time series. Otherwise, the flux correlates well 
with the ring current. The vertical beam position drifts down likely due to the
monochromator stabilizing. It takes about 5 hours for the monochromator to 
stabilize. Again, the beam moves slightly horizontally after the fill on fill.

Fig 02-01.6. Time series of the beam intensity in 7ID-C, starting at 16h30 on 02/01/04 and
lasting over 10 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is not connected.

Fig 02-01.7. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 02-01.6.

Monday Feb 2, 2004:summary of start up operation by Don Walko.

A final time serie was started on Monday morning, at 6h13 am. The intensities,
and beam positions are shown in the next two figures. Note again the large 
noise on the vertical beam position.

Fig 02-02.1. Time series of the beam intensity in 7ID-C, starting at 06h13 on 02/02/04 and
lasting over 3 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is not connected.

Fig 02-02.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 02-02.1.

Summary included in the TRR group meeting minutes of 02/02/04:
Don Walko reported on various improvements or alignment corrections which
have been made to the 7ID beamline.  The APS Survey Group, in particular
Scott Wesling, found that the Commissioning Window in 7ID-A was not quite
in place, and moved the upstream end outboard by 0.018".  The White Beam
slit tank is at its correct location.  Weasel also found that the
micromono tank in 7ID-B may have been slightly tilted, with the upstream
end low by 0.015" and the downstream end high by 0.018", but on average
was at the correct height.  The tilt did not significantly change when the
micromono tank was moved inboard on its rails; and this tilt was not
corrected.  The micromono was then opened and a Compton shield was shimmed
higher, which seems to have removed a blockage.

The cryocooler for the 7ID-A HHL monochromator has been operating better
since its last annual service: a leak on the high-pressure side was
repaired, so the vessel will no longer need to be filled up every week or
two; and the low-pressure fill modification installed last year was
removed, reducing the pressure spikes and presumably reducing beam motion.

Don also reported that when Dohn Arms upgraded the EPICS installation of
the 7ida crate, that caused problems with the horizontal-blade motors of
the white-beam slits.  These motors have brakes which were apparently
always turned off in the old EPICS.  The best solution seems to be to
leave the brakes off, although further tests are needed to be sure the
slit motions are reproducible.  We should try to get them serviced in the
near future.

Wednesday Feb 4, 2004: Lining up the HHLM tank horizontally and vertically.

Fig. 02-04.1 shows the white beam intensity transmitted
through the Si (111) thin wedge as a function of the horizontal tank position. 
The mono tank was moved by 1.26 mm to make sure that the thin part of the mono
was centered on the undulator axis. This alignment was not done when ED first 
installed the first crystal in the tank in late January 2001. This realignment
may reduce the power absorbed by the first crystal. The first crystal angle was
20 degrees.

Fig. 02-04.1: Recentering the table of the High Heat Load monochromator horizontally.

Fig. 02-04.2 shows the white beam intensity transmitted
through the Si (111) thin wedge as a function of the Bragg angle. Above 8.75 
degree, the transmitted intensity could be controlled by the angle. One would
expect a well aligned first crystal would be able to control the transmitted 
intensity as low as zero grazing angle of incidence. This scan revealed that the
monochromator table was too low. The tank was raised twice, once y2 was raised 
manually, and once y1 and y3 were raised.

Fig. 02-04.2: transmission versus angle with table jack too low.

Fig. 02-04.3 shows the white beam intensity transmitted
versus the angle once the table was raised in its final position. The intensity
for Kamel Fezzaa can now be controlled to 4 degree. We need to repair the 
broken monochromator table jack and realign completely the monochromator.

Fig. 02-04.3: transmission versus angle with table jack at the final position.

Feb. 11, 2004: Energy calibration of mono after moving the tank up.

On February 11, Don Walko recalibrated the HHLM on the Zr edge for B. Adams's
experiment. Near the Zr edge at 18 keV, the mono was off 19 eV because we moved 
the monochromator table jack y2 manually on Feb. 4 for Kamel Fezzaa's white beam
experiments using the thin crystal of the monochromator as a white beam filter.

Fig. 02-11.1 shows the Zr edge calibrated scan. The first, second, and third
bumps on the scans are respectively 18.001, 18.014, 18.031 keV. A follow up scan
near the Nb edge is also shown (around 18.991 keV) in Fig. 02-11.2. One keV away
from the Zr edge, the monochromator calibration was found to be off by 3-5 eV,
which might indicate a poor alignment of the monochromator.

A long time serie started on 2/11 at 17h35 is shown in Fig. 02-11.3 and
Fig. 02-11.4. The APS ran in top-up mode, the monochromator was set to
18 keV, the 7ID energy was set to n=3, E ~ 18.1 keV.The intensities in the 7ID-B
ion chamber and the 7ID-C diode track well each other. Several small tweaks of
the second crystal may have happened during the time series.

The vertical beam position is quite stable over this 11 day long period once the
temperature of the monochromator has settled down. A long drift is noticable 
with a motion on the order of 10 microns. The large high frequency noise has
been seen earlier in the run and will be investigated in future studies. The
horizontal beam position is less stable. 

Fig. 02-11: Recalibration of High Heat Load monochromator energy after a change
of the y2 table height at the Zr edge.

Fig. 02-11.2: Calibration scan near the Nb edge.

Fig 02-11.3. Time series of the beam intensity in 7ID-C, starting at 17h35 on 02/11/04 and
lasting over 275 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is not valid as the signal is saturated.

Fig 02-11.4. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 02-11.4.

Tuesday Mar 2, 2004: New EPICS screen by Dohn Arms.

Fig. 03-02: a new EPICS screen showing the hutch and floor temperature and pressure.

Fig 03-02 shows a new EPICS screen was made by Dohn Arms on 3/02/04. It shows
the temperature of the experimental all and the barometric pressure.  Note that
AC has been installed in 7ID-D and the sensor is inside the hutch. Eric Landahl 
and Dohn Arms are working to reduce the set point to make it colder inside the 
hutch.  All other sensors are over the hutches or over our heads. The label 
number of the sensor i.e. 5045 is shown on the AC duct.

Mar. 10, 2004: Shielding validation of 7ID-D, surveying one last time the beam line and Time series.

Due to making a new labyrinth on top of the 7ID-D roof for the AC input, the
7ID-D had a routine shielding validation today at 11 am and passed without 
problems.

Don Walko called S. Weasels  and he came around 2 pm to survey the horizontal
placement of the white beam. The L5-20 were first centered with the 
monochromatic beam. He found no significant offset. The beamline is ready to
install a small polished Be WI-82 window in 7ID-A. (to be continued)

Mar. 11, 2004: Time series and investigation of the noise on the BPM vertical position.

Today, we are investigating the increased noise on the vertical beam position
signal. Fig. 03-11.1 and Fig. 03-11.2 show a short time serie started at 
3h37 am, with the monochromator set at 10 keV, the L5-20 slit opening set to
0.75 mm (H) by 0.5 mm (V), and the ID gap set to 10.055 keV.

Fig. 03-11.3 to Fig. 03-11.5 show a short time serie started at 13h32 before
the TRR group meeting. Note the correlation between beam motion and LN2
cryocooler vessel fill.

Fig. 03-11.6 shows the proper calculation for getting Volts into degree C 
for the second crystal thermometer transducer. Do not touch this record.

Fig. 03-11.7 to Fig. 03-11.9 show a 6 hour long time serie started at 18h41 before
the TRR group meeting. Note the correlation between beam motion and LN2
cryocooler vessel fill.

Fig 03-11.1. Time series of the beam intensity in 7ID-C, starting at 03h37 on 03/11/04 and
lasting about 0.62 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The effect of top-up is clearly seen.

Fig 03-11.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-11.1. The standard deviation of the vertical
beam position is about 3.5 times larger than the one for the horizontal position. Note the bump
seen in the 7ID-C ion chamber near t=0.25 h correlates to a small vertical upward motion.

Fig 03-11.3. Time series of the beam intensity in 7ID-C, starting at 13h32 on 03/11/04 and
lasting about 2 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is covered by several filters so one reads dark noise.

Fig 03-11.4. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-11.3. The horizontal and vertical motion
(starting at t = 0.25 h) during the fill are respectively 40 and 25 microns. The motion is correlated
to the cryocooler fill of the low pressure vessel(see next figure).

Fig 03-11.5. The cryocooler level sensor. The bottom plot is broken.
Clearly the fill induces beam motion.

Fig 03-11.6. I fixed the broken calculation record that converts the 2nd crystal temperature transducers
units (V) into degree C. Please do not touch this user calculation until a permanent fix is completed.

Fig 03-11.7. Time series of the beam intensity in 7ID-C, starting at 18h41 on 03/11/04 and
lasting about 6.1 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C ion chamber is covered by several filters so one reads dark noise.
There was a 7ID-B hutch access near t = 5 h.

Fig 03-11.8. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-11.7.

Fig 03-11.9. The cryocooler level sensor. The bottom plot has been fixed towards the end of the time serie.

Mar. 12, 2004: Several time series and effect of the cryocooler pressure set point on beam position!

A time series were started just before bed. Fig. 03-12.1, 03-12.2, 
and 03-12.3 show the data from the time series.
Note that the temperature of the second crystal is now also recorded.
The monochromator is set to 10 keV, 7ID to 10.055 keV on n=1, and the white beam
slits are set to 0.5 mm (V) by 0.75 mm (H).

Fig 03-12.1. Time series of the beam intensity in 7ID-C, starting at 00h29 on 03/12/04 and
lasting about 4.5 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The 7ID-C chamber is not biased properly.

Fig 03-12.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-12.1.

Fig 03-12.3. The cryocooler level sensor and the second crystal temperature. The 2nd crystal temperature is
often noisy and hard to interpret.

Fig 03-12.4. Time series of the beam intensity in 7ID-C, starting at 17h45 on 03/12/04 and
lasting about 4.6 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.

A time series were started just before dinner. Fig. 03-12.4, 03-12.5, 
and 03-12.6 show the data from the time series.

Fig 03-12.5. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-12.4.

Fig 03-12.6. The cryocooler level sensor and the second crystal temperature.

I have characterized the first crystal mount sensitivity to a pressure change
by pressurizing the closed loop without at room temperature. I never did 
carefully measure the pressure dependence of the beam position. Tonight I
pressurized the closed loop in steps of 1 PSI and recorded a time serie. 
Fig. 03-12.7, 03-12.8, and 03-12.9 show the effect 
of set point pressure on the beam position. The effect is easy to observe. Table
03-12.1 shows the pressure steps. The X position has a sensitivity of 10 microns
per PSI, while the Y position moves downwards with slope of 5 microns/PSI.

Table 03-12.1
--------------------------------------------------
time (h)     Pressure (PSI)
--------------------------------------------------
0-0.19         15
0.19-.392      16
0.392-0.625    17
0.625-0.861    18
0.861-1.055    19
1.055          start of cryocooler fill
--------------------------------------------------

Fig 03-12.7. Time series of the beam intensity in 7ID-C, starting at 23h54 on 03/12/04 and
lasting about 1.1 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.

Fig 03-12.8. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-12.4.

Fig 03-12.9. The cryocooler level sensor and the second crystal temperature.

Mar. 13, 2004: Several time series, recentering of the X-ray BPM.

A time series were started just before bed. Fig. 03-13.1, 03-13.2, 
and 03-13.3 show the data from the time series.
The monochromator is set to 10 keV, 7ID to 10.055 keV on n=1, and the white beam
slits are set to 0.5 mm (V) by 0.75 mm (H). 

Fig 03-13.1. Time series of the beam intensity in 7ID-C, starting at 01h52 on 03/13/04 and
lasting about 11.3 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.

Fig 03-13.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-13.1.
The was a LN2 fill at the end of the time serie, and the pressure in the crocooler closed
loop rose to 22.3 PSI from 20 PSI in a few minutes.

Fig 03-13.3. The cryocooler level sensor and the second crystal temperature.

Mar. 14, 2004: Several time series, in particular the atmospheric pressure and the first crystal temperature.

In the past 40 h, I took a long time serie to see whether the suggestion that 
Bob Dortwegt made in his Feb. 16 2004 email would translate into observable changes.
Fig. 03-14.1 and 03-14.2 show the first crystal temperature and the 
atmospheric pressure changes in a 40 h long time series started on 3/12/2004 at
22h17. My temperature sensor is too noisy to look at very subtle and long 
drifts. No significant correlated change is observed on T1 for a .0135 bar
variation in 40 h. Perhaps someone has a more sensitive sensor and can further 
investigate this! Bob suggested a 1/15 bar change (0.066 bar) would result in 
a 0.5 C changes in LN2 temperature. The atmospheric pressure change observed 
here is 1/5 of the one suggested by Bob, thus one would have expected subtle
temperature drifts here of 0.1 C.

Fig 03-14.1. The monochromator first crystal temperature in a long time series.
The data from this sensor is very noisy and it uses only 12 bit digitization.
A 300s running average (30 pts) was performed to improve the quality of the data.
The temperature baseline appears to drift down by about 0.1 C.

Fig 03-14.2. The atmospheric pressure in the same time serie as Fig 03-14.1.
The pressure variation are about 0.013 bar. How much would this affect the boiling point of LN2?

Time series started on 3/13/04 at 16h44

Fig. 03-14.3, 03-14.4, and 03-14.5 show a long time serie started at 16h44,
with the monochromator set at 10 keV, the L5-20 slit opening set to 0.75 mm (H)
by 0.5 mm (V), and the ID gap set to 10.055 keV. The figure captions are self 
explanatory.

Fig 03-13.3. Time series of the beam intensity in 7ID-C, starting at 16h44 on 03/13/04 and
lasting about 28 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
The beam was tweaked around t = 24h, and several hutch access interrupt the time series.
The slit in front of the ion chamber in 7ID-C must clip the beam after the tweak.

Fig 03-14.4. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-13.3. After the tweak, the beam moved
up by about 0.3 mm.

Fig 03-14.5. The cryocooler level sensor and the second crystal temperature.

Evening of 3/14/04: Repair and recalibrated the Y BPM stage

Tonight, I also fixed the Y BPM EPICS setting so that the motion is accurate
and calibrated. I set up a digital dial gauge, our Mitutoyo that was used on the
monochromator table in February. I reduced the speed, increased the backlash 
(major point), and increased the acceleration time. I also set up the rest and 
idle to off on the Next Step driver. Before these changes, the slide would not 
move properly. Afterwards, the slide motion is off by 2 microns per mm of
travel!  I recalibrated by scanning the BPM Y position +/- 1 mm and fitted the
curve of YRatio to a line. The calibration was hardly changed. I noticed that 
over the range of the calibration, there is some obvious non-linear S shape 
turns.

Tuesday, March 16, 2004: Main beamline IOC failure and reboot today due to a network connection lost.

Today Dohn Arms had to reboot the main beamline crate due to network problems.
It appears that the failure occured at 2h49 am on 3/16/04 and was repaired at 
17h42 today.(See Fig. 03-16.1) We had problems to restore the monochromatic beam. The 
piezo voltage was still at 5.4 V after the reboot but the piezo at the first be 
tweaked at 6.12 V then finally optimized at 5.1V. So it appears that one may 
need to take into account the hysteresis loop when the  crate goes down since 
the autosave and restoring process cannot take care of hysteresis.

On Sunday 3/14, I started a time serie at 22h29, and the beam position is shown
in Fig. 03-16.2. Note that I turned off the BPM Y motor 
driver after I recalibrated the axis. It is interesting to not the long 5 hour
long drift that follows turning off the driver. I increased the backlash on the
slide to improve the motion accuracy, so perhaps when one turns the power off, 
the slide slowly creeps. There is evidence in the plot for 60 um motion in 5 
hours. I do not believe the beam moved significantly during this series.

Fig 03-16.1. The strip chart recording of T1 and Patms. The time when the T1 measurement is
interrupted is caused by the network failure.

Fig 03-16.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during a 25 hour long time serie.

Tuesday, March 17, 2004: 7ID-C hutch IOC failure.

Today, I had to reboot the 7ID-C hutch IOC. The scan record would not work
properly. Once I noticed that the scan record was not working, the IOC crashed.
It was rebooted. At 3h20 pm, I also retweaked the monochromator piezo voltage 
from 5.1 V to its original 5.6 V so the heating of the monochromator resulted 
in a beam drift of 240 um, and a flux reduction of 0.5/4.5 = 11 % drop. This
is noticable in Fig. 03-17.1 and Fig. 03-17.2 (see vertical motion).
The monochromator tweak is clearly seen in the beam position time series.
Today, I also borrowed a pressure regulator and plan to purchase a pressure
monitoring system to control the pressure of the high pressure vessel.

Fig 03-17.1. Time series of the beam intensity in 7ID-C, starting at 13h04 on 03/17/04 and
lasting about 68 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
There are several 7ID-C hutch access that interrupts the time series.

Fig 03-17.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-17.1.

Sunday, March 21, 2004: time series.

Fig. 03-21.1 and 03-21.2 show a time serie started at 00h35 on 3/21 and 
lasting 10 hours. The 7ID Gap was 21.7613 mm, the 7ID monochromator Energy was
10 keV, and the white beam slits were set to 0.75 mm (H) by 0.5 mm (V).
The average beam position is stable to about 5 microns over this 10 hours time
period, if one ignores the beam motion correlated to LN2 fills.

Fig 03-21.1. Time series of the beam intensity in 7ID-C, starting at 00h35 on 03/21/04 and
lasting about 10 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
A 7ID-C hutch access interrupts the time series near t = 8.7h. The 7ID-C ion chamber signal is
saturated.

Fig 03-21.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-21.1.

Wednesday, March 24, 2004: realignment of L5-20.

Today, ED realigned the L5-20 and tweaked the monochromator. Yesterday, we also
moved the small optical table from the wet lab into the 7ID-B hutch yesterday, 
and lowered the 7ID-C exit table for C. Schroer's camera.
Fig. 03-24.1 and 03-24.2 show the intensities and beam positions 
for a short time series started at 14h48 on 3/24. Fast horizontal beam position
changes are noticable with a fast onset of 5 um motion and a slow relaxation. 
This is caused by the sharp turning on and off of the heater element. 


A beamline EPS tripped the beam last night around 11pm, closing GV5 for no
apparent reasons. I received a page from Christian, but was fast asleep at the
time. This morning, the gate valve was reopened and the beam is back. We have no
idea why the valve closed.  The BLEPS pager alarm for notification of BLEPS
problem affecting PS2 has been implemented. ED received a page when the APS
went down yesterday, but did not receive one when 7ID went down.

On Thursday night 3/25, the same EPS trip occured around 11h20pm. 

Fig 03-24.1. Time series of the beam intensity in 7ID-C, starting at 14h48 on 03/24/04 and
lasting about 400 seconds. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.

Fig 03-24.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load
mono in 7ID-C during the same time series as Fig. 03-24.1.

Sunday, March 28, 2004: problems with the cryocooler buffer level, running out of He in 7ID-B.

Sunday, March 28, 2003:
ED came around 19h30 to check on the cryocooler buffer level. On Friday night,
it was around 60% before he left. Tonight it was 51.5%. So perhaps when ED 
serviced the jackets last Tuesday, one of the valves was not sealed properly and
the vacuum seal went bad afterwards. We will have to take care of this Tuesday
i.e. check what caused the LN2 to leak. Note that the level has not be filled
once this cycle in contrast to previous runs where we would fill the buffer 
every two weeks.

ED also replaced the He bottle in 7ID-B. It had completely ran out when he 
catched it. We should design some safeguard so that it does not happen again.
Perhaps a double He bottle farm would improve this and a sensor with an alarm
telling us when it runs out. Fortunately we are running at 19 keV, so the ozone
generation was probably lower than it would be typically  at lower energy.
The resulting damage to the window was probably reduced due to this choice of 
energy. 

Monday, March 29: The cryolevel is now 49%(14h40), so it dropped 2.5% in 20 hours.
We should be OK to run until tomorrow morning. We definitely need to check what
is going on tomorrow first thing in the morning.

Tuesday, March 30, 2004: Filling the crycooler buffer, pumping the jacket and new pressure control.

Today, the crycooler level problem was investigated. The high pressure buffer
was filled with LN2, the cryocooler jackets attached to the mono tank were 
pumped and the ED installed a new pressure regulator to control the high 
pressure buffer pressure. The new circuit controls the pressure better than 0.1
PSI, probably to a few 0.01 PSI. This will result in submicron beam stability.
It also keeps the closed loop filled becuase the dry nitrogen from the boil off 
is condensed in the buffer. I watched the buffer level rise today by about 2%
in 1.5 hours!

Thursday, April 1, 2004: time series analysis.

Fig. 04-01.1, 04-01.2, and Fig. 04-01.3 show a 2.8 hours long time series
with the monochromator set to diffract 15.0 keV X-rays. The white beam slits
were set to their typical value of 0.5 mm by 0.5 mm and 7ID (n=3) = 15.1 keV.

Note the horizontal beam position (the vertical position has noise we suspect
is caused by bad electronics). After the fill, it stabilizes rapidly and an 
average from 0.2 h to 0.5 h has a standard deviation of 0.7 um RMS. Recall that
in the past it had 5 microns amplitude fluctuations (see previous data).

The pressure regulator provides a continuous supply of nitrogen gas that
condenses in the buffer, thus keeping the buffer full at 101.8% from now on. The
presssure excursion during the fill have been reduced by a factor 3, and the 
pressure regulator seems to bleed the extra gas generated by warming the buffer
during the fill.

Fig 04-01.1. Time series of the beam intensity in 7ID-C, starting at 12h37 on 04/01/04 and
lasting about 2.8 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
Several hutch access interrupt the time series.

Fig 04-01.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load mono
in 7ID-C during the same time series as Fig. 04-01.1. At t = 1.5h, Don Walko accessed 7ID-B
to change the nitrogen bottle. It is not clear why the beam position changes by 8 microns following
the hutch access.

Fig 04-01.3. The cryocooler level sensor and the second crystal temperature during the same time series. The fill completes around t = 0.15h.

Monday, April 5, 2004: Solenoid replacement.

Today, while checking the new regulator circuit, ED noticed that the low 
pressure Dewar was not automatically filling anymore. The Dewar level was at 36
%, instead of being between 50-60%. The APS system was functioning fine, so we
found out that the cryogenic fill valve controlled by a solenoid was not
responding anymore. The solenoid was checked for continuity and found to be
broken in an open circuit mode instead of a few hundred Ohms.  We were lucky to
borrow a spare from sector 6. We only replaced the solenoid. The system is now
working, and no beam time was lost. (See Fig. 4-05.4 and Fig. 4-05.5.)

Fig. 4-05.1 show a 2.8 day long time serie started on Thursday April 1, at 12h37 pm.
The monochromator was set to diffract 15.0 keV X-rays. The white beam slits
were set to their typical value of 0.5 mm by 0.5 mm and 7ID (n=3) = 15.1 keV.
Note that top-up failed on Saturday near t = 42 hrs and Sunday near 66 hrs.
The intensity of the 7ID-B ion chamber and 7ID-C diode correlate with this ring
current drop.

Fig. 4-05.2 show the beam position during the same time series as Fig. 4-05.1.
Fig. 4-05.3 shows also the LN2 level sensor varying from 50 to 60% every 4 hours. 
Note that the rapid beam motion occuring during the fill is no longer obvious to
see (compare Fig. 4-05.2 with Fig. 3-21.2).

Fig 04-05.1. Time series of the beam intensity in 7ID-C, starting at 12h37 on 04/01/04 and
lasting about 2.8 days. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.
Several hutch access interrupt the time series. Note that top up was interrupted near t = 42h
on Saturday 4/3/04 and at the end of the time serie near t=66 hours.

Fig 04-05.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load mono
in 7ID-C during the same time series as Fig. 04-05.1. The large horizontal and vertical beam motion correlate
to a top-up failure.

Fig 04-05.3. The cryocooler level sensor and the second crystal temperature during the same time series.

Fig 04-05.4. The broken solenoid coil.

Fig 04-05.4. The new solenoid coil that was replaced in 7ID-A.

Tuesday, April 6, 2004.

Today, I fixed a leak in the new pressure regulator circuit. A festo connector
was leaking when it was getting cold. Because of the pressure bump during a fill, 
cold nitrogen gas flows back to the regulator where it leaks out. The festo
connector was getting cold because of this. I replaced the festo connector with
a 3/8" Swagelock connector.

I also fixed an EPS problem. It turns out that when we brought in the ladder in
7ID-A yesterday, the thermocouple input of the first crystal temperature on the
mono detached from the Acromag cold cathode compensator. I noticed it first 
because the first crystal temperature was 50C even though the Si block is cold.
Tonight I noticed the yellow cable was detached from the Acromag Intellipack 
801T-1500 so I reconnected it. This would have tripped the EPS system at 8am.

Tuesday, April 9, 2004.

Fig. 04-09.1, Fig. 04-09.2, and Fig. 04-09.3 show a time series started on 04/09
and lasting about 10 hours. The monochromator is set at 10 keV, 7ID at 10.054 
keV, and the white beam slits are 0.5mm x 0.5 mm wide. The beam position is 
is stable to about 5 microns. Note the small 5 micron horizontal motion 
remaining during the cryocooler fill. 

Fig. 04-09.4 shows the beam current per bucket in the 1+7x8 hybrid mode. The 
singlet current is 8 mA.

Fig 04-09.1. Time series of the beam intensity in 7ID-C, starting at 01h23 on 04/09/04 and
lasting about 10 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.

Fig 04-09.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load mono
in 7ID-C during the same time series as Fig. 04-09.1.

Fig 04-09.3. The cryocooler level sensor and the second crystal temperature during the same time series.

Fig 04-09.4. The beam curent per bucket on 04/09/04.

Monday, April 12, 2004. Short time serie and streak camera data.

Fig. 04-12.1, Fig. 04-12.2, and Fig. 04-12.3 show a time series started on 04/12
and lasting about 0.8 hours. The monochromator is set at 10 keV, 7ID at 10.054 
keV, and the white beam slits are 0.5mm x 0.5 mm wide. The beam position is 
is stable to about 10 microns. Note the small 10 micron horizontal motion 
remaining during the cryocooler fill. The beam may have been a bit more stable
when I first tried this trick because the leak in the festo connector acted as 
a pressure relief valve. Since I fixed the leak by replacing the festo connector
with a Swagelock connector, I've noticed that the pressure bump is higher than 
it was during the first week of tests. See the work done on last Tuesday.

Fig. 04-12.4 and Fig. 04-12.5 show data obtained from B. Wang at S35 on single bunch mode.

Fig 04-12.1. Time series of the beam intensity in 7ID-C, starting at 11h51 on 04/12/04 and
lasting about 0.8 hours. The 7ID-C diode sum, the ring current, ion chamber in B and C are shown.

Fig 04-12.2. The beam position, 49.2 m from the source, or 19 m from the High Heat Load mono
in 7ID-C during the same time series as Fig. 04-12.1.

Fig 04-12.3. The cryocooler level sensor and the second crystal temperature during the same time series.

Fig 04-12.4. The streak camera data from S35 BM on 04/12/04.

Fig 04-12.5. The streak camera Bunch 0 data from S35 BM on 04/12/04.

Lessons learned during run 2 of FY04

By removing the cryocooler low pressure fill modification, we have found that 
the Cryocooler fill pressure disturbance are less severe, and more rapid to 
return. The beam position variations are also less severe.


To do list:

-check and fix Y-BPM motor/driver (see 1/29/03). (fixed! 3/14/04 by ED)
-complete remote shutter interface
-Reduce the speed of the pressure response of the Huber chamber gauge.
-Prevent the turbo from venting when the pump is turned off or a power failure
 occurs.