MHATT-CAT redesign of our ID-line
Monochromator First crystal mount
Eric Dufresne, MHATT-CAT, January 2001
Table of Content:
Motivation for the upgrade and acknowledgements.
The work shown below is being done to improve the stability of our LN2 cooled
HHL Monochromator. We have found in the last year that we experience beam
motion on the order of 100 micron in hutches 7ID-C and 7ID-D due to pressure
fluctuations which induce angular tilts of the first crystal. For details,
please follow this link.. Starting last May, a collaboration led by Yizhak
Yacoby from Hebrew University designed and build a new first crystal mount to
stabilize the first crystal. The parts were received in early August, but
several modifications had to be made to install it in the monochromator. Some
were made in August, the rest of them through the last fall. Several parts had
to be procured for example such as bellows which only arrived in late November
The following MHATT-CATers have been involved in this project: Yizhak Yacoby,
Ron Pindak, Bob Macharrie, John Pitney, Steve Dierker, Roy Clarke, Dohn Arms
and myself. The latest modifications are described below. I wish to thank the
technical staff at Hebrew University, at the APS machine shop (Bruce Hoster,
John Wozniak, Bob Conley), the APS Optics Fabrication Lab., the Central ANL
shop (Bill Toter, Al Schneider, Wes, Krysty, Bob), and the MHATT-CAT
administrative staff (Cynthia D'Agostino and Charles Sutton) at the University
of Michigan for their hard work.
Projected schedule for Installation, 1/8/2001
After talking with the shop on Friday, it appears that the shop will not
be too busy next week. This is encouraging. The optics shop needs to do
the manifold polishing, but I don't know how busy they are yet. This is
something that will be clearer tomorrow. Before I start jobs in the shop,
I will likely need tomorrow and Tuesday to make and/or modify drawings.
The following parts as defined in Yizhak Powerpoint drawings need
work (see following figure):
part 1: (Invar manifold) First weld small pipes to invar manifold to be
able to butt weld the bellows afterwards. (1 day) Then polish the manifold
on the sealing side after the pipes have been welded to it (2-3 days) Job
start 01/10 and completes by 01/12 4pm
part 2: (Manifold holder with dove tail) It needs a 45 degree millout portion
because it hits the sine bar. It needs also a small cut out to make sure
it clears the pump lip. Also the G10 M3 tapped holes need to be redrilled
45 degrees from the existing hole to match the clearance hole made on part
1. Also the dove tail needs to be refinished since its surface was scratched
(one day). Job start 01/10 and completes by 01/11 4pm
part 3: The dove tail surface needs repolishing. Also the hole for the
rotation pin needs to be plugged and drilled again for a smaller pin. It
must be milled also to clear the roller holder. Job start 01/10 and completes
by 01/11 4pm
part 4,4a: remachine so that the roller holder do not hit these plates.
the rotation pin as well the plates need to be milled I need to also open
up a few virtual leaks in these plates (2 days in parallel with part 1,5)
Job start 01/10 and completes by 01/12 4pm
part 5: (block to LN2 lines) Remake in invar and redesign the part so that
the LN2 pipes connect to it directly on the same side as the existing manifold.
(2 days in parallel with part 1, likely done in another ANL shop). Job
start 01/10 and completes by 01/12 4pm
pipes: Complete design and weld to part 1 and to part 5, and to the 1.33"
flanges. (2 days) Job start 01/15 and completes by 01/16 4pm
assemble: assemble everything after UHV cleaning, make a new seal, align
in the tank, and do some pressure test with tilt sensors. (1 day) Job start
01/17 and completes by 01/17 end of day
cool: pump (typically 1 day, but could take longer if leaks are found)
and cool the optics (1/2 day) and start testing. Job start 01/17 and completes
by 01/18 end of day
I would expect to be ready for the first cooling test optimistically by
the beginning of the run, not quite on the 17 but shortly after. I doubt
things will go this smoothly though. If the mono is cooled successfully,
I would benefit from some help to do stability measurements by the weekend
of the 19th of January. I would think we could repeat Steve's measurements
with a quadrant diode to test the stability again in ID-D. Also Dohn will
start by the 20th or so thus I'll have more help later on. If we need a
second pass, then help will be welcome to debug further.
Wednesday January 10, 2001
Late Tuesday night, I finished the majority of the drawings (modifications
to part 1,2,3,4,4a, and the redesign of the manifold part 5) and early
yesterday I put jobs in the shop. Unfortunately, I missed my "one" day
window of opportunity. The shop is now busy and it does not look like they'll
be able to make it by Friday. They sometimes work overtime on Saturday.
I'll speak to the shop manager later on today to see what they can do.
I do believe part 1 was sent outside for welding though so this
is good. Once I get it back, it can be polished by the optics group. They'll
leak test it too because making an invar to stainless weld is supposed
to be quite tricky. So this is an additional test I did not know about.
Thursday January 11, 2001
I do hope to get most of the machining done on the parts submitted yesterday
by the end of the day Monday. They started to work on it today. The stainless
to invar weld was completed and leak tested this afternoon in the main
ANL shop and the polishing work will be queued in tomorrow morning in the
optics shop. I spoke to a machinist there and it is very unlikely that
the job will start tomorrow, so we can't expect this part back before at
the earliest Monday afternoon.
Since the elbows have been ordered and should arrive tomorrow,
you might be aware that I specified some nice 90 degree elbows to connect
the piping together. They are short, but designed to have a nice smooth
radius on their corner. They will be butt welded to 1/2" OD pipe. I have
another order I'll send in soon for special tools for helping with operational
issues in the mono installation i.e. long hex keys, etc...
So, I should have all the machined parts finished hopefully by
Tuesday for assembly. I might have to wait a little for the invar seal
polishing, so this will be clearer tomorrow or Monday. I still have to
complete the pipe design tomorrow and then I'll be able to start the welding
of the flanges to 1/2" pipes etc... This should speed up the final assembly
of the bellows and pipe.
Friday, January 12, 2001
Some good progress was made by the shop today. They finished the machining
of part 2,3,4,4a and delivered it to me. So I can make sure that the mechanical
issues have all been resolved by installing it in the monochromator chamber.
I'll do this this weekend. Part 1's polishing was started today and will
be finished Monday. Part 5 slipped from my attention today (remaking of
the invar manifold). I'll have to find out what is going on with it Monday.
There was a lead time in getting invar so I am not sure if the job started.
It will also need welding afterwards. I received the 90 degrees elbow.
We have the bellows in hand.
Sunday, January 14, 2001
I came in and tested all the parts received Friday afternoon. They all
fit nicely in the chamber and the mono goes to 41 degrees easily. The back
plates no longer interfere with the motion of the second crystal and in
principle one could go to 3 keV on Si (111) with this crystal if the second
crystal was long enough to intercept the beam... The monochromatic beam is not
at a fixed offset for Bragg angles above 32 degrees. We are still roughly on
schedule. I also prepared a few small modifications I had forgotten to do last
week, i.e. the dove tail was sticking a bit, so I asked for a gentle repolish.
I also requested a refitting of the new pin because it did not allow for plate
3 and 4 to be in contact. Finally, I vented 14 possible virtual leaks on
plate 4a. These are quick job I'll submit Monday. Below are some pictures
Above, plates 4 and 4a were shortened to clear the roller holder bracket
which is near the second crystal Cu holder.
Above, plates 3, 4 and 4a and dove tail bracket (part 2) installed in the
chamber. Note the 45 degree milled out part that allows to go to as much as
Same as before, at 41 degrees. The sine bar would have hit around 29 degrees.
Note the bottom right corner of part 2 was cut.
Same at zero degree. Note that two slots near the G10 post recess were added
to help removing part 1 when we need to make a new crystal seal.
Above, closer look at part 3 and 4. Note the pin diameter has been reduced to
10 mm, and all the plates length were cut by 14 mm. Without this cut, the
second crystal would not move properly to keep a fixed offset. Note that the
hole in plate 3 had to be plugged and redrilled to 10 mm.
Monday, January 15, 2001
This morning, I deposited several jobs in the shop and requested an update on
the new manifold (part 5), made in invar. Bruce told me the job is waiting for
Invar, and after checking with his purchasing agent, Invar won't come in until
Friday 1-19. Of course, this is a big delay, so I walked around the APS and
ended up visiting Wah-Keat Lee from SRI-1 who gave me some. I took it from Dale
Ferguson in 401-L0119. So I took the invar to the shop, and in the afternoon,
they were able to start the work. Also, the polishing of the sealing manifold
(part 1) was completed. It took them 0.5 hour for grinding, 1 hour for fine
grinding and finally 6.5 hours to polish down to 1 micron. So, about 8 hours of
shop for making a good LN2 seal. It is well worth it for a good LN2 seal.
By the early afternoon, all the parts submitted in the morning were completed
and everything looks great. At night, the pipe design was completed.
If everything goes well, tomorrow, we should start welding all the piping.
Tuesday, January 16, 2001
On the good news, Charles found the missing MDC bellow so he'll ship
it to me tomorrow, so in case we have problems with welding, we have spares.
I also have another spare borrowed from PNC so we are OK.
Part 5 was completed and I received it by mid morning (see figure below). The
shop made it in two parts, a 10 mm thick plate and a manifold, tied together
by two bolts. So I started to prepare everything for welding, but I soon
realized that it would greatly help the assembly process to machine the plate
further to have better access to the G10 posts screws and to the bottom dove
tail slide set screw. The manifolds makes these screws hard if not impossible
to access. Also I found out the screws that tie the manifold needed to be
vented. So I drew some modifications, submitted the job to the shop and got it
back by 4h pm. In the meantime, I worked on a sheet metal shield for protecting
the polished Invar surface for welding.
At night, I worked to prepare the job. I assembled all the flanges, bellows,
seals, etc to be ready for dropping parts in the ANL central shop in the
morning. I rechecked the pipe drawings and design. I also collected all
the large parts that need UHV cleaning. Preparing for welding was a bigger
task than anticipated since final machining had to be done and delays with
fabrication of part 5 occurred. Find below annotated pictures of the day's work.
Dove tail assembly and rotation plates after repolishing on Monday.
Above, Newly machined part 5, LN2 manifold.
Above, Monday, 7 vent holes we drilled in plate 4a (3 are seen here on
Above, On Monday, 7 additional holes were slotted in plate 4a.
Above, Tuesday, initial assembly of part 2 and 5 in preparation for welding.
Same as above, note access problem to G10 screws and slide bottom
Above, pre-assembly with part 1,2,5 shown.
Above, modifications to part 5 received Tuesday afternoon. Note the new
and clearance holes.
Same as previous image, other side shown. The two screws were also
Wednesday 1am, everything is ready for welding. The shield is on.
Wednesday, January 17, 2001
This morning, I took the parts for welding and UHV cleaning to Bldg 363. I met
with Bill Toter (2-7342). He could not start immediately because the pipes
were not made yet. I did not get the Stainless Steel pipes done earlier.
He did not know when he could deliver initially. In the afternoon, he
confirmed that the welding should be completed tomorrow afternoon with
possible use of overtime. The pipes should be ready tomorrow morning. So we
should be ready to assemble late tomorrow. Looking back at the schedule, we are
about a day late. The parts have been UHV cleaned so I can pick them up
tomorrow. I received a call late yesterday and the shop needed some feedback.
They suggested to put the conflat not off-center on the pipe connecting the
elbows because it might block the side G10 posts, so we agreed to put the
flanges slightly below the center of the pipe. They had a good idea. I had to
walk to Bldg 363, but I also received some of the the cleaned parts. Tomorrow
I can start to work on the back plates installation.
Thursday, January 18, 2001
In preparation for installation later today, I started to clean bolts and find
all the required hardware to make the In seal. It took me a while to find the
spare kit, and worried for a while, but I found everything to make a seal.
Al Schneider called me and we agreed I would come over around 2h30 once they
leak checked all the parts. I wanted to discuss cleaning the remaining parts
with them. I went there around 2h30 and witnessed how they clean the parts.
They do a good job. So by late afternoon, I had all the parts in hand UHV
At night, John and I did a large fraction of the installation but since I was
not able to get a telescope to align the unit today, I thought it was not
worth killing ourselves tonight. We had to do last minute machining (cut a
vented M6 bolt, file 2 of the G10 posts so that all 6 posts can be in, make a
tool to access the Huber screws). I think it might be better to have 6 G10 posts
installed instead of 4. I decided to use the spare crystal with the new unit
because I want to keep the old unit intact to have as a backup. The spare
crystal surface is not as nice, and last May I had strain problems with it,
but we traced these to machanical strain from the pipe retaining screw. The
topography work showed this crystal was fine.
Thursday night, the robot as John calls it is pieced together. Note the
Closer look at the pipe, with its bellow and elbow.
The bellow should work nicely since it is
stiff longitudinally but easy to bend transversely, thus helping with the
The other manifold (part 5) with its new piping.
Friday, January 19, 2001
Today, John and I assembled the remaining items. We made a new seal, then put
the dove tail on its mating unit. We ran into an immediate problem with it
i.e. the two part nearly got totally binded. Fortunately, I was able to take
them apart. We needed a lubricant for the dove tail and for this matter the
rotation pin. After a search, Wah-Keat Lee gave me a lubricant that has been
used on the UHV Kohzu monochromators called MORESCO-GK1. It has low vapor
pressure and is radiation hardened. We should buy some. So with lubricant,
everything worked smoothly. The assembly of the roation stage was no fun.
Two of its three retaining screws are nearly impossible to access. The bottom
right one was particularly difficult to insert. One the unit was in place,
we attached the conflats, levelled the Huber Horizontal slide, levelled the
crystal. We found in the end a chi offset of 0.6 degrees, the inboard side
being higher. We levelled the crystal height with a telescope and set the
1st and second crystal vertical offset.
The pressure test revealed some great news. The new unit flex only by 0.2
microradian/PSI compared to 9.5 microradian/PSI with the old unit. This is an
improvement of a factor 47.5.
The assembled unit. The LN2 hose are connected and are about 1cm above
where they use to be on
the old design. It was hard to attach the conflats
that are in between the bellows due to the relatively
easy torsion and
flex of the bellows.
The assembled unit again, a left view showing the pipes.
The assembled unit again, showing the manifold.
The assembled unit again, right view. Note that we installed all the
G10 posts in.
Saturday, January 20, 2001
Today, we pumped the chamber while sleeping, then turned the ion pumps on
around 4h30 pm. We started also the cryocooler to cool the pump and fill the
Dewar. At night, we started to cool the optics but unfortunately a huge leak
developped when the temperature of the crystal reached roughly -130 C. It was
so big of a leak that LN2 started to leak on the bottom of the tank. The ion
pump tripped in advance but I got confused because the back Huber ion pump
first noticed the leak. Fortunately, I had loosened all the tank bolts so the
door opened slightly to leak the N2 gas. We have to leak check tonight once the
temperature of the crystal is warm enough. We filled the tank with dry N2
to further limit water condensation.
After opening, because we had such a large leak, we decided to bubble ethanol
on the crystal assembly. It does seem like ethanol was bubbling on the crystal
seal so I retigthened it quite a bit. We were able to pressurize the lines
without any leaks (a leak would have shown some gas flow). Thus we pumped the
chamber and leak checked with an APS leak detector. No leaks were found when
flowing He in the pipes. We called it a night and are pumping the chamber
Sunday, January 21, 2001
Today, we were successful in cooling the optics. No leak whatsoever occurred
so everything is going well. The new design of the first crystal mount is now
installed in the beamline and has seen light. Tonight, we roughly aligned the
two crystal using an ion chamber in air placed in 7ID-B and slowly tuning until
the beam appeared. It worked quite well. We brought beam as well in 7ID-C. Our
first preliminary news is that the beam is now not lost at all during a fill,
i.e no major dip of intensity anymore. We see some large intensity fluctuations
but we have to do a more careful alignment first and close the slits too. With
a 3mm by 2mm beam, we observed 20 % fluctuation in intensity peak to peak. This
is not too good, but then again, this is a large white beam on the
monochromator so the test conditions are quite different from the previous one
i.e. 0.4 by 0.4 mm.
Monday, January 22, 2001
This morning I received a relatively agressive message from Didier at sector 6
who complained about the noise of our cryocooler. The problem is that last
December I found I had an Oxygen deficiency problem in 7ID-A because I was
venting the cryocooler inside the hutch. The APS requested we stop doing this,
so I did on Sunday start to vent outside with the result that it generated again
a huge amount of noise like it used to.
I talked to Didier and promised to do something about it. In the afternoon,
I found a way to muffle the sound. It does look like it is an end effect, i.e.
a lot of Nitrogen gas going through a narrow orifice. I taped a garden
hose at the end of the pipe that the laser group bought and it worked fine.
On the beamline front, we've had several problems with our shutter, especially
the P4. The Love controller on it keeps on tripping over a limit, so the
APS enable is removed. This now has happened at least three times in
24 hours, the first time yesterday, and at least twice today. During one
occasion, the P2 (2nd photon shutter) also tripped but this was quickly fixed.
Due to several distractions (muzzling, P4), it took a while to get started with
the mono alignment. I found the total flux to be reasonable at the moment,
although it still seems a little low, i.e. 6.6x10^12 ph/s/100mA.
I'll have to do a good alignment of the mono.
I started by aligning the L5-20 in both direction, the vertical slit being off
by 0.5 mm. Then I roughly realigned the mono table height which was off by 3 mm.
I went to white beam operation late in the evening to align the first crystal
angle but I had a hard time interpreting what I was measuring. I could not
find the zero of theta easily in a plot of intensity versus theta. I called it
Tuesday, January 23, 2001
Today the beam was down from 8 am to 4 pm, so in the evening, I realigned the
first crystal in the beam carefully. I set an ion chamber in He in 7ID-B and
found out a large first crystal misalignment. The surface was off in theta by
0.51 degrees, which explains my problems of yesterday. The crystal was leveled
with respect to the beam by looking at the reflected beam from the Si surface.
Then, the beam height was readjusted (I had problem with hhly2 which lost its
position and sounds rough when moved), then the horizontal position was aligned
with the first crystal at 20 degrees to line up the thin web in the beam. I had
to fix the hhlx1 and x2 motor record a little, and noticed that x1 receives
a few random steps per minute, so it is best to leave these two motors off.
I then calibrated the mono energy with the Zn edge.
So on the positive side, the total flux is not lost anymore during a LN2 fill!
Because of the changes I made yesterday for muzzling the sound, this creates
some back pressure in the vent line which seems to warm up and pressurize the
closed loop, by up to 4.7 PSI! This is a major improvements.
I did not notice any serious improvements in scanning the mono in energy. There
are still significant time dependent intensity fluctuations with a peak to peak
variation of 4 % approximately with an 85 seconds period or so. A slow drift
needs to be investigated as well. Note that the white beam slits are quite
widely open: we typically operate with 0.4 mm by 0.4 mm.
On the annoying side, possibly due to an alignment problem, I don't have the
same flux we use to have, being perhaps a factor 3 lower than expected. It
could be also due to the fact that I had to retighten the crystal seal
significantly straining the crystal. Find below a time serie of the total flux.
The total monochromatic flux (times 10^12 ph/s/100mA) with a 3mm by 1 mm
white beam in 7ID-A.
Note the slow decay by 2.5 % over 45 minutes. This
will need to be investigated further since
I seem to recall the old design was pretty insensitive to long drifts.
Also note the fast fluctuations
on a time scale of 20/14 minutes or
about 85 seconds. These fast fluctuations could be the BESSRC
Elliptical Multipole Wiggler.
Wednesday, January 24, 2001
Today I started to plan to install a quadrant beam position monitor in 7ID-C
to monitor the total flux and beam position in that hutch. While doing this,
I also took data thus find below a few representative plots. It turns out that
the monochromatic flux is fairly stable over ten hours once corrections for beam
decay are taken. Last night, some decay was seen likely due to some heating
on the first crystal of the monochromator. I retweaked the second crystal today
and gained 25 % more flux. More importantly tonight, I confirmed with the APS
that the 85 second sinusoidal fluctuations are indeed the Multipole Wiggler.
Our monochromatic flux should be far more stable if the Elliptical Wiggler
was off. Roy was not impressed by that when we found out... To make matter
worse, the feedback is off because the software is broken. It caused beam dumps
at the beginning of this run. So we may have to request the device to be off
to take good data...
The total monochromatic flux (times 10^12 ph/s/100mA) with a 3 mm by 1 mm
after a retweak of
the monochromator second crystal. Note the effect of the
fill on fill of the APS at 8 pm.
The beam is stable to about 5.0 +/- 0.2 (4%)
in the figure. After the fill on fill, some relaxation
seems to occur on the order of one hour. The
data starts shortly after the tweak of the second
The total monochromatic flux (times 10^12 ph/s/100mA) with a 3 mm by 1 mm,
for the last hour.
The average flux is quite stable over an hour. It is
confirmed: the 85 second fluctuations are due
to the BESSRC Wiggler.
Thursday, January 25, 2001
Today, the Elliptical Wiggler was turned off until tomorrow morning. The figure
below shows that the periodic fluctuations disappear, although some small
flux fluctuations remain. During several hutch access, the P4 was closed by
mistake explaining the sudden drops. Tomorrow morning, the EMW should be pulsed
again. Tonight we set up a quadrant diode detector at 50.5 m from the source,
i.e 20.5 m from the High heat load mono. The electronics is hooked up and some
rough alignment scans were performed tonight.
The total monochromatic flux (times 10^12 ph/s/100mA) with a 3 mm by 1 mm,
for the last hour on
1/26/2001 with the BESSRC Wiggler OFF!
Note that the 85 seconds fluctuations are gone.
Several closure of the P4
caused the flux to vanish temporarily between -25 and -45 minutes.
Friday, January 26, 2001
The BESSERC Wiggler is on again, so are the fluctuations (see figure below).
Today, we aligned the chi of the second crystal to send the beam in ID-C close
to the ideal position. I would hope the beam will be within a mm of the path
the white beam would have followed.
Dohn and I played further with the quadrant diode detector and got it centered.
I had to leave to take care of my wife who caught a bad case of the flu. The
test will hopefully resume tomorrow.
The total monochromatic flux (times 10^12 ph/s/100mA) with a 3 mm by 1 mm,
for the last hour on
1/25/2001 with the BESSRC Wiggler ON again...
Saturday, January 27, 2001
I came in to install the last flight path in 7ID-C, then set up the electronics
for the quadrant diode. The diodes are each connected to an SR 570, each set
to 20nA/V, with a -12dB low pass 10Hz filter. At 100 mA, each diodes measures
about 3 V, and without beam the rms fluctuations range between 0.26-0.47 mV,
which is at about the single bit level of the digitizer(0.305mV). From standard
propagation of error, assuming Gaussian noise, the rms fluctuations on the beam
positions are expected to be (with 50 mA in the machine) 1.3 micron (Hor) and
0.9 micron (Vert), this being the worst case scenario. The digitization step
results also in approximately 1 micron resolution. Two figures are shown below.
The first figure shows the sum of the 4 diodes versus time showing the same type
of fluctuations detected with the ion chamber. Although not shown, the signal
from the ion chamber in 7ID-B and the sum signal in 7ID-C correlate perfectly.
The fluctuations are about 3 % peak to peak (p2p). I confirmed that the EMW
was on but the amplitude seems smaller than earlier on this week. I closed
the white beam slits yesterday, but I would suspect this makes matter worst.
Perhaps the APS has tried to improve the beam stability since yesterday.
The second figure shows the beam position versus time. We see large horizontal
motion of about 15 um p2p, and a smaller vertical motion of 10 um p2p. The dark
heavy line is a 15 second running average of the X position. The beam motion
in X and Y show the same periodicity of 87 seconds or so. It would be nice to
ask BESSRC to turn off their Wiggler temporarily to test our beam stability.
A 10 um stability in beam position represents a 10um/20.5m/2 ~ .25 urad
stability in Bragg angle. Without the Wiggler on, we should be able to do
By the way, all this data was taken from home (thank God for cable modem)...
Tomorrow if I can, I may take longer time series to see whether the LN2 fill
is observable, and whether long drifts are present. Also, I may look at the
flux versus energy curves.
The sum of 4 diodes with a 2 mm by 0.6 mm white beam on the mono with
the BESSRC Wiggler ON.
The beam position in 7ID-C with the BESSRC Wiggler ON. The heavy line
is a running
average of the X position.
Sunday, January 28, 2001
Sunday, the beam was down for quite a while. Following a question from Steve
Dierker with regards to noise issues, I estimated the noise one would expect
from counting statistics alone. Using the fluorescent yield of Ti of about 12%,
our solid angle fraction of 1.4% for a single diode, and estimating the flux
in 7ID-C to be 10^12 ph/s photon on the Ti foil, each diodes collect about
1.7E9 photons/s, assuming unit quantum efficiency of the diode(overestimate
here). One would think the relative noise would be sqrt(1/1.7e9)= 24ppm,
thus on a 1.5 V signal, 0.04 mV, which is well below the electronic noise...
It is probably not it, but maybe there are other source of noise, possibly from
elastic scattering from the windows for example.
There was an afternoon dump around 1h30 PM. Below you'll find the long Sunday
overnight time serie started at 2h11 am. Note the LN2 fill causes small beam
motion separated by 4 hours or so. Long drifts are seen with beam motion of
roughly 100 um in both directions. (See 02/01 for the horizontal drift cause.)
The beam dumped Sunday night at 10 pm, so I had to wait for a while to get
started again. In the meantime, I found out that the fastest I could drive
the 16 bit A/D at the moment was about 1 Hz, so I decided to change the filter
frequency on the SR 570 to 1 Hz. This change did not in the end make much of
a difference on the signal. I programmed a time serie collection C-routine using
EZCA to get all the normalization and hopefully get the fastest sampling
possible. Once the beam came back, due to the large beam motion seen since
I first align, I had to recenter the BPM in the horizontal by 600 microns
and took the opportunity to do a better calibration curve.
One graphs shown below starts at 3h01 am and emphasize the short time scales.
Two graphs are shown next taken after 4h05 am. The short time motion is on
order of 10 microns. Drifts specially in the horizontal are large, but their
origin is not clear at this point (see 2/01 for details).
the IDC driver, moving the stage in one direction). Do they come from parts
in the mono heating up or from the lack of feedback on the APS?
The LN2 fill causes small positional motion because the pressure rises by 4
PSI during a fill, causing a 4psi x 0.2urad/psi = 0.8 urad shift in the first
crystal. One would expect the beam to move by 2x0.8urad*20.6m = 33 micron
during a LN2 fill. The vertical motion observed is actually only 15 microns.
The beam position in 7ID-C on Sunday morning 1/28. The time serie starts
at 2h11 am and lasts for
6.7 hours, starting just after a machine fill.
Note the beam motion bumps separated by 4 hours.
The beam position in 7ID-C after the APS feedforward was implemented by
Louis Emery. The time
serie is started at 3h01 am and lasts for 200 s,
to emphasize the shorter time fluctuations.
The beam position in 7ID-C after the APS feedforward was implemented by
Louis Emery. The time serie
is started at 4h05 am and lasts for 1000 s.
The vertical (red) and horizontal (black) beam motion are
10 and 15 micron respectively peak to peak.
The beam position in 7ID-C after the APS feedforward was implemented by
Louis Emery, starting
at 4h05 am and lasting until the beam dump.
Note the large horizontal drift observed of more than 200
The bump in X (44 micron) and Y (15 micron) at 9000 seconds is due to the LN2
fill of the
Wednesday, January 31, 2001
Today, I realigned the white beam slits and peaked the undulator gap for the
mono Bragg angle set to 10.0 keV. The white beam slits were only off by -119 (V)
and 87.4 (H) micron thus we can't blame large horizontal drifts on the APS. The
beam hasn't move more than 100 micron in a week. The total flux is off maybe
a factor 2.5 so we have about 4.2x10^12 ph/s/100mA. The horizontal drifts
observed with the BPM must be caused by our set up, likely beam motion
(see 02-01 for details).
The total flux is a bit more stable than it was before (see below). 90 seconds
fluctuations are still present, but their amplitude is smaller (3.6 % p2p) than
they were last Friday for example, where they were possibly twice as large.
The EMW is still on.
The total flux in 7ID-B after the APS feedback was turned on.
The total flux in 7ID-B after the APS feedback was turned on. The second
crystal was retweaked
after 2h40min hours of flux decay and has been
stable for the last hour and a half or so. Thermal drifts
might be present which take several hours to equilibrate. I would recommend we
at all time on the monochromator during this run.
Thursday, February 01, 2001: Conclusions.
Please note the following errata from yesterday. The horizontal drifts observed
for last week were actually caused by faulty stepper motor grounding or by
noise pickup from the power lines. The IDC Next Step drivers were receiving
noise on its stepping line and actually stepping the motors. This problem is
known and was fixed today by Dohn Arms and I. Dohn and I grounded the input
cable of the IDC on motor cable 301-312. So please ignore the large horizontal
drifts seen earlier in the BPM data. The BPM stage was moving not the beam!
(Also see figure below for corrected data on the BPM.)
(See this page on 9/6/2000 for details.)
Please note also that on 2/15/2001 we moved the beam 40 microradians down
and we finally cleared an edge that made us more sensitive to the EMW
I would conclude this commentary by saying essentially that the mono is more
stable now than it was before to pressure fluctuations. The slow drifts that
were observed were bugs, so we'll have to experience with this set up more to
see how it behaves. The fast time stucture could not be analyzed to better than
one Hz. The stability limitation at the moment is caused by the APS EMW. The LN2
fill is far less of an effect that it was before, and it is hard to observe on
the flux alone. Improvements (factor 2 reduction? ) can be made hopefully this
run. Today, we looked at the beam with a YAG/video camera and found a huge
improvement in beam stability as compared to the old design. The team work done
on this project was well worth it! Thanks to all who were involved.
This concludes this web page. Actually not... see the P.S. for the latest!
The Corrected beam position in 7ID-C after the APS feedforward was
implemented by Louis Emery,
starting at 4h05 am and lasting until the beam
dump. I subtracted a linear drift to the X position data.
The bump in X (50 micron) and Y (15 micron) at 9000 seconds is due to
the LN2 fill of the Oxford cryocooler.
Monday, February 19, 2001: Post Scriptum.
P.S. Note that the intensity fluctuations that were seen on January 2001 have
pretty much disappeared following an orbit correction performed on 2/15/2001.
During the last alignment of the mono, some mistakes were made (possibly the
offset is a little too large) which thus sent the beam a fraction of a mm to
high. After the micromono exit Be window in the 7ID-B hutch, the beam had a
sharp horizontal edge blocking the top part of the beam. We've known for a
while but we did not have any diagnostics in the beamline until last December
2000. On 2/15, we looked at the beam in 7ID-A just after the mono and found it
was not appertured at all. So it is certain that the edge is either caused by
the P4, or an aperture in the micromono (a Be window maybe?). This will have
to be investigated further in March 2001.
Because the edge cut off the beam on axis so much, what I aligned on was
a low energy wing of the undulator spectrum. It was shown by Sandy et al.
(JSR (1999) 6, p1174.) that the beamprofile at an energy about 0.5 keV below
a 7.66 keV fundamental at the APS has a double peak profile with less intensity
right on the undulator axis. What I did turn out to do was just to align our
first crystal at 10 keV and peak the total integrated flux at 10.8 keV thus
the bottom peak on the double peak profile was not apertured and it peaked
up for a gap of 10.8 keV. The horizontal edge really blocked the on-axis
flux. This explains the poorer flux that was seen by 7ID-C experimenters during
their experiment and likely the low gain seen with the new mirror.
On 2/15, Dohn Arms, Steve Dierker and I performed a 40 microradian orbit
steering correction and we were able to clear the aperture. We now have a
10 keV X-ray peak in the flux with a gap set to 10.05 keV (like before),
the beam is on axis. The total flux is now as much as 2.5E13 ph/s/100mA, which
is about a facor 2 better than the best we've done before. So this is good news
for future experiments with the HHL monochromator.
Below is a time serie of the total flux in 7ID-B. The total flux is now stable
to about 0.35 % peak to peak over one hour. Note that the time fluctuations of
the total intensity still have some periodicity, but they have improved by an
order of magnitude almost by resteering the beam. The beam is now far more
stable than it has ever been.
The total 10 keV flux measured in 7ID-B with the white beam slits set to
2.5 mm (V) by 1 mm (H).
The total flux is limited by the L5-20 to
1.7E13 ph/s/100mA here and the peak to peak fluxtuations
are only 0.35 %!
We are now far less sensitive to the EMW than in January, and indeed the APS
improved their orbit stability implemented by Louis Emery. The LN2 cryocooler
filled 20 minutes
before the end of the time series showing that we are
now pretty insensitive to it.