US workers at GS: Chuck B. (NSH reporter), Fabrizio C., Lori G.,
Erik K., Ioannis K., Sophia K., Massimo O., Paulo S., Stefano S.
(Stephane C. gave you their last names. I give you their first names
since I have trouble spelling names with more than 40 characters...)
Table of Contents:
A) Abstract.
B) MACRO RUN summary.
C) MAINTENANCE OPERATIONS.
D) Appendix on "Afterpulsing Revisited".
E) Rich Baker Memorial Quiz.
For those of you who like to read only the abstract of papers:
A) ABSTRACT: Everything at MACRO is running smoothly. Period.
(Now for a long and boring diatribe. You may want to stop here. Don't
say I didn't warn you. I'm going for the LONGEST SHIFT REPORT record.
I believe in quantity, and (obviously) not quality.)
B) MACRO RUN summary:
In the 13 days I was in Italy, there was only one unplanned downtime
for the detector. On Thursday 27 November (that's Thanksgiving for you
gluttonous Americans) a PHRASE power supply gave up the ghost between
8:00 and 9:00 just as Stefano was riding the Navetta into the tunnel.
He replaced it and Erik's excellent WWW trigger rate monitor (check it
out at the MACRO Online web page:
wsgs02.infn.it:8000/macro )
showed that we only lost about 1 hour of data taking. On the weekend prior
to my arrival (on Saturday 15 Nov) there was a period of very inefficient
data taking due to an overactive thyroid in a TOHM trigger requesting too
much WFD readout. Unfortunately (but all too familiar to electronics techs)
the problem turned out to be intermittant and went away before the source
of the spazm could be found. I had the thrill of trying to learn PASCAL
and making a modification to the DAQ software to minimize the impact of
such a convulsion in the future. (See more on that below.)
On Saturday 22 Nov while an ignorant American "weekender" was watching
things, a Streamer Tube power supply tripped (five minutes after the Italian
shift monitors left for Italian lunch). After calling every number I could
find scribbled on the counting house wall (including the one that read:
"for a good ____job, call..." and arranging an evening rendesvous), I finally
got ahold of Erik who informed me that this was a recurrent (but again,
annoyingly intermittant) problem and that sitting tight and letting things
run was the best course. Dekhissi reset the supply upon his return.
C) MAINTENANCE OPERATIONS.
I will now summarize some of the activities which went on while
I was present at GS. Although it may look like I did the most work,
that is grossly misleading. The paid workers (and Sophia, too) went about
their work diligently and efficiently while I played around with some of the
EXCELLENT analysis software that geniuses too numerous to mention (no, not
you, Greg) have written over the years. It's just that the part I did is
about the only thing I understand, and as you will see, even this is an
overstatement.
I) LeCroy HV module problems.
Paolo S. and Massimo O.
Since the power problems initially reported by MACRO's sole Nittany Lion
collaborator, the 32 V power supply of the LeCroy HV system of SM 5 has
been acting up. It seems to be particularly sensitive to input voltage
fluctuations. The only spare module resident at GS has the SAME problem.
The modules of SM1 and SM5 were swapped (and later unswapped), but the
bug migrated with the swap, implicating the 32 V p.s.
The resident electronics gurus suspect that the "beer can" electrolytic
capacitors in the 32V supplies are dying, causing the problems. Massimo
is supposed to be contacting TAMU for further instructions.
II) ERP S/H module replacement.
Ioannis K. and Chuck B.
On Wednesday 19 Nov we swapped out one erp S/H modules in SM2:
2W01-04 : The ADC-A vs. ADC-D plot presented a very "fuzzy" picture. It
seems that either ADC-As or ADC-Ds have some noise component.
On Wednesday 26 Nov we swapped out three erp S/H modules in SM4:
4B01 : The ADC-A vs. ADC-D plot for side 1 presented a parabola instead
of the typical line with slope ~=10
4W07 : The ADC-A vs. ADC-D plot for side (? don't remember) is a line
with slope = 2 instead of the typical ~=10.
4T10 : The ADC-A and ADC-D info are inconsistent; large jitter.
After these changes, calibrations (laser and LED) indicated that these
channels and their friends and neighbors (i.e. part of the same module)
were working.
The following erp channel is still under suspicion, but has not been
replaced:
6S05 : Were not able to troubleshoot it, but the LED and LASER code
are unable to calibrate. Ioannis is planning on swapping this
module out this week (starting 1 Dec) in time for Wednesday's
cal run.
III) LED problems.
Stefano S. and Ioannis K.
On Wed 19 Nov Ioannis reported the following:
"Having fixed almost all standing LED problems on SM1 and SM2, we
decided to look into the only one remaining, that of 2T07-1.
This showed as a dead LED for both LED/TDC and LaMOSSKa calibrations.
As it turns out, the LED cable for this tank-end is broken. We have
tested both LEDs and they seem to be fine. In fact, the 'North' side
LED (the one closer to 2T01-10) has a slightly higher light output.
We left the LED cable unplugged for 2T07-1. It should be taken
care off by the end of this week."
"Looking at the LED/TDC cal. data we got on 19 Nov. for SM1 and SM2, it was
found that the switchbox for 1C01-08, side-0 must be broken: all TDC signals
for these 8 counters present a very high jitter (8-12ticks), and all for
side-0. If IOK remembers well, during LED/FMT, he also noticed these counters
missing. LaMOSSKa should definitely pick them up, also.
The other problems, still remaining are:
1E02-0 -> we guess this is a low-light LED
2T07-1 -> bad LED cable.
The plan is to fix these before the next SM1 & SM2 cal."(10 Dec?)
On Wed 26 Nov SS and IK reported the following:
a) 4b10-0 was found to have a bad BNC cable (a bad BNC connector
on the switchbox end). The cable was replaced.
b) 3b07-1 and 4b02-0 show high end jitter (> 10 erp TDC ticks). This
problem is small enough that no change in hardware is planned for now.
c) SM 3 & SM4 in particular are in good shape for ERP calibration.
d) For the entire MACRO, the current ERP TDC cal efficiency is better
than 99.8% (i.e. 99.8% of channels are giving good TDC data, and the
other 0.2% are mostly from bad cal hardware, not bad ERP channels).
IV) TOHM calibration procedure (under development; not yet incororated
into standard calibration procedures)
Lori G. and Ioannis K.
This procedure does a forced (i.e. non-triggered) readout of the WFD's.
A software bug was discovered, and when fixed, a successful test was
performed on SM 3&4. A more thorough test will be performed on Wed
3 Dec.
V) Exxon Valdez cleanup...
Massimo O. and Stefano S.
An end chamber leak in tank 1b03 (forgot which end, but it is a tank
which has a retrofitted PVC end chamber) was drained, opened, inpsected,
had it's o-ring cleaned (VERY IMPORTANT!), closed and refilled. After
a few days it appears as though this tank end is no longer leaking.
One styrofoam coffee cup is now freed up for Greek coffee drinking (the
mineral oil can't help but improve the taste; anyone got any spare
pseudocumene?). Current goal is one tank end leak fix per week. There
are currently on the order of 20 of these, all minor and all pure oil,
I think.
VI) Spaztic TOHMs.
Erik K. and Chuck B.
As mentioned above in the "MACRO RUN Summary" section, the weekend
before I arrived the data taking rate went through the roof due to a
spaztic TOHM trigger (sorry, but I forgot the tank end; Stefano or Erik
should be able to fill in the blanks). After consulting Ed K. and Chris O.
at BU, Erik assigned me the task of writing a software "fix" (not really)
which will bypass WFD readout caused by TOHM triggers from a noisy tank end.
(Note that because the problem hasn't recurred, it is not known what its
source is--could be a sparking tube??)
Basically an array is set up in the WFD_patreg routine which keeps
track of TOHM triggers as a function of tank end. When a MAX LIMIT (to
be entered via the CAMAC list) is reached for a given tank end, any future
TOHM request to have that particular tank end read out will be ignored.
This array will be zeroed at the beginning of a run (restarting the
procedure). We were successful in turning off channels, but unfortunately
too successful--we also turned off the Siamese channel in the neighboring
SM. The torch will be passed to a future unsuspecting shift worker (whose
initials might just be Dan Levin) for more robust PASCAL coding.
And some very brief reports on other MACRO related work being carried out:
VII) TOHM-LI interface testing has begun. Stefano S.
VIII) Laser splitter boxes are being surveyed for problems. Also SM1
plasma cartridge is being replaced before 3 Dec. (NOTE: I believe this
is the last spare plasma cartridge, so whoever is responsible may want
to get some more ordered.) Stefano S.
IX) Paolo S. continues his odyssey or converting the "Wrong List" to a user
friendly Web page. When he's finished, you may want to direct all your
C++ questions in his direction.
X) The latest "Wolverine in residence" Fabrizio C. is working on combining
PHRASE and ERP data to do a multiface search for fast (>10^3 c) monopoles.
D) Appendix. Afterpulsing revisited.
(Congratulations if you've made it this far. You will be given extra
credit on the quiz which follows.)
Besides spending time on the DAQ code, I also was given the opportunity
to look at the waveform data on Erik's "RARE DST's". I am molto impressed
(sincerely). There is a goldmine of data anxiously awaiting physicists to
start searching for exotic events. Just scanning some of these using all
the nice DASH and DASH++ software was a thrill. In particular, I spent
quite a bit of time on three bizarre events. One was an anomolous PHRASE
event which Fabrizio C. reported on at the recent collab meeting and which
Erik has done considerable further study (which can be found on his WEB
site--URL mentioned much earlier. You'll need to enter as MACRO with
the standard password since he doesn't want some snotnosed hacker-snoop
getting access.) With the help of Erik, Fabrizio, and Ioannis, we were able
to determine the nature of two of the three events. The third is still a
mystery.
In the process of looking at the PHRASE event, I recalled some work
done by Steve Ahlen and Alex Marin (see Macro Internal Note 1022/91 and
also the minutes of the Bologna meeting from the fall of 1991) where
they proposed using afterpulseheight analysis to extend the dynamic
range of the PM's. They also proposed using afterpulseheight analysis
to monitor the helium contamination of the PM's. I did some investigation
on both fronts and report the results here.
Basically, the afterpulses they (and I) are referring to are those
caused by gas atoms/molecules inside the PM. When the electrons from the
photocathode get enough energy (near the first dynode) they can ionize
any ambient gas inside the PM. These ions then drift back to the
photocathode (with a characteristic time proportional to molecular wt ^-0.5)
and further ionize the photocathode, giving a secondary pulse. Ahlen and
Marin studied H, D, and He "spectral lines".
In the PHRASE event, one tank end showed a hugely saturated PM
signal, so I used muon WFD signals (which don't saturate the anode) and
associated afterpulses to calibrate the He-afterpulse height for that
particular tank end (2C14-1). Those tubes showed a He-afterpulse to
primary pulse ratio of about 1/50. (Note that in the small sample survey
which follows, this number is considerably lower than average.) The
He-afterpulse height for this tank end for the PHRASE event was 2.4 V!
Thus the "unsaturated" primary would be in the neighborhood of 120 V. It
turns out that an E-M shower from a muon occurred DIRECTLY in front of
the PM, where the light production has a 1/x^2 component (x being the
distance from the scintillation production point to the PM).
The second proposal of Ahlen and Marin was to monitor the ratio
of primary pulseheight to He-afterpulse height as a function of time
to see if the PM's would eventually die due to He contamination (from
leaks in the streamer tube gas system). I don't know if a followup
was ever done (until now). In fact, they measured and reported the
He afterpulse height for tank ends 2b15-0 and 2b15-1. Although it
is not 100% clear from their work, they report a hydrogen to helium
pulse height of 9::1 for what I believe is the -1 end. (My reasoning
comes from the fact that the WFD data shows an unusually prominent
H line for this tank end.) I measured the same ratio at about 2::1.
Taking their idea (but modifying it slightly and comparing the
H- to He-afterpulse heights vs. time), then assuming no change in
H content of the tube, the He content has gone up by a factor of
roughly 4 or 5 over a time period of 6 1/2 years.
According to Stefano, there are certain PM's which show considerably
worse He-afterpulsing than others. It is speculated that these PM's
are closest to streamer tube chamber gas leaks. For a random sample
of tank end hits in supermodule 2, 40 measurements of the primary (voltage)
pulse height to He-afterpulse height ratio had extrema of 30 and 300.
(Note that errors on a single measurement are on the order of 30% for my
crude methods.) The mean for these 40 measurements was 104 and the
standard deviation was 61.
E) Rich Baker Memorial Quiz. (No, he's not dead.)
One thing I've missed the last couple trips to GS is sitting at
the Quatro Specchi and playing Rich Baker word games with Rich and other
MACROites. In reminiscing, I've come up with the following quiz. Although
not nearly as sophisticated or internally consistent as one of Rich's
inventions, nor as clever as what Bob and Chris have come up with, this
is my tribute to MACRO on what is likely my last shift. The idea is
to use the crossword-puzzle-like clues to come up with names of persons,
places, and things dear (or not so dear) to MACROites. Here is an example.
wealthy cyclist (= rich biker == Rich Baker). Note that once you have the
initial words translated, you often still end up with a STRONG accent or
two. (Can you hear the Australian accent here? Other potential accents
are Italian, US southern, US northeastern, Hoosier...) Note below that a
dbl--hyphen indicates that the two clue words are to be translated
separately (as opposed to the two words going together making
a single idea ==> syllable or name.) The clues are indicated E, M, or D
for what I consider easy, moderate, and difficult.
Category 1: Things. Identify the clues with MACRO objects (e.g. parts
of the detector).
Milton Bradley board game (E)
Wild West's Wyatt (E)
MACRO USA spokesman (E)
a heavy, learned volume (E)
(ok, now they get tougher)
Queen's utterance when knighting the Trashbag: "I ___ ___ _____" (D)
Durocher's nickname (M)
arrived--US dog truck--fewest (M)
bovine pie plate--cgs magnetic unit (M)
opposite of fuse (M)
Russian mutton dish (M)
Category 2: Places. Identify the clues with locations familiar to MACROites.
(hint: All but 2 are less than 20 km from the detector. Of the remaining,
one is in Italy and the other is a US institution.)
woman--region--honeymaker (M)
Norwegian Texas fare (D)
mom's mom's anus (M)
not many appliance--negative (E)
MACRO's Hanson--straw mattress (M)
shortage of weeping trees (M)
dad left near crow's cry (D)
maize shuck--Hillbilly matron (D)
ore tester--follows "golly!" (M)
Category 3: Mammals. Match the clue in column A with a creature from
column B. (L ==> last name, F ==> first name, otherwise full name).
hint: All are proper names.
precious jewel (E) Italian spokesman (L)
pessimistic, investmentwise (E) Bologna's Steve (L)
MACRO's Lori obtained a Hog (D) MACRO file overseer
the bend in a drill wrench (D) Frascati K-9
fold--sheep cloth--rip (M) US spokesman (L)
flag of Italian apples (M) Mediteranean tavoli player (L)
sticky money (D) fluffy towel man
Durocher cheer--roast Oscar Robertson (D) champion jetsetter
Cassius Clay's wood finishing tool (E) geologist's son
liked reject of mine (D) beantown Abruzzesi
masticated MACRO's Spencer (E) AIX jock
admired sister's felines (D) Minos Minor
Bronx cheer for US national anthem writer (M) wife of one of the above (F)
small hint: "Trashbag", "Hog", and "Hillbilly" are capitalized for a reason
Any spelling errors are unintentional! ("Minor" is not mispelled, but
pun intended.)
Send entries to Rich Baker--rbaker@msc.cornell.edu
Prize for perfect score: dinner for four at Quatro Specchi (airfare not
included).