==> Next meetings: wednesday, july, 15th, 1998, 10:00 h
tuesday, sept., 1st, 1998, 9:00 h
Present: B. Allongue CERN
G. Bianchetti CERN
J. Bohm PRAGUE
G. Dumont CERN
J. Elias FERMILAB
P. Farthouat CERN
Z. Hajduk CERN & CRACOW
B. Hallgren CERN
R. Hammarstrom CERN
M. Hass CERN & WEIZMANN
P. Jarron CERN
N. Lupu TECHNION
E. Radermacher CERN
R. Richter CERN & MPI
W. v. Sprolant CERN
G. Stefanini CERN
1. R. Richter: Practical aspects and schedule for n-irrad. at PROSPERO
The n-dose in 10 years of LHC is predicted for CMS to be about 3 * 10^10
n/cm^2 in the central region (eta=0) and about a factor 10 more in the
forward region. In ATLAS both numbers are a factor of 10 higher, respectively.
In a first n-irrad. run we would like to scan a broad range
of n-doses by selecting doses in steps of 10 (e.g. 3*10^10, 3*10^11,
3*10^12 and 10^13).
The PROSPERO facility, mentionned in the last minutes, provides n-doses
of up to 10^14 n/cm^2 per hour. The spatial distribution is following
roughly the 1/r^2-law (not perfectly, because of n-reflection from
the walls). Practical distances from the reactor-axis are r = 0.2 to 5
m.
At 4 m distance from the axis of the reactor one obtains 3*10^11 n/cm^2
per hour at 100% reactor power (3 kW). These are 1 MeV equiv. neutrons
w.r.t.
damage induced in Silicon. One can operate at 10% reactor power and
get 3*10^10 n/cm^2 in 1 hour, the 10-years dose for CMS (central region).
Because of the large distance of 4 m, there is plenty of space for
equipment in the otherwise empty hall. During the same period of 1 hour,
at
r = 1 m, one obtains about 10 times more, which corresponds to the
ATLAS 10-years dose in the central region.
Some test devices should be removed from the test area after the first
hour of low-power irradiation in order to study these devices for long-term
anealing effects later on. This interruption, together with the necessary
cool-down time for personal access, will take less than 1 hour.
After this one can go to 100% reactor power for 3 hours and accumulate
3*10^12 n/cm^2 at r = 2 m and 10^13 at r = 1 m.- The total test lasts
about 5 hours, covering all 4 above-mentionned doses. Other doses can,
of course, be selected at different distances.
We foresee sept., 30th as the date for the first irrad. run. People
from the LHC community, who want to participate in this run should give
notice a.s.a.p. and specify requirements of space and doses. It is
understood that every user has to prepare all equipment (cabling, power
supplies, R/O) for his test on his own. Also installation at the reactor
and monitoring during the irrad. are the responsibility of the user. As
mentionned before, cables must be at least 20 m long to service equipment
in the test area from the control room. (Please remember that power
connectors are different at CERN (swiss standard) and in France.)
People who want to participate in the test, must send their personal
coordinates plus a copy of there passport to the center in the first days
of september in order to get clearance for the site. The site is located
about 30 km north of Dijon.
2. G. Dumont: Irrad. tests of DC/DC converters at the ESS group
The electronics pool will participate in the n-irrad. tests of DC-DC converters. The following topics will be addressed:
- survey the market
- collect information on rad-tol components
- define the electrical specs for units to be tested (e.g. power)
- select specific DC-DC converters and buy items for test
- build up a set-up for rad-tol tests
- participate in pre- and post-rad tests
- evaluate global cost of test, incl. manpower
Particular responsibilities have been taken:
Market survey:
G. Bianchetti, B. Allongue
Rad-tol matter: B.
Allongue
Magn. field matter: W. v. Sprolant
Follow-up:
G. Dumont
3. B. Hallgren: Preliminary Results from irrad. in TCC2
Bjorn described the TCC2 radiation area in LAB II, where the accelerator
division has started to test CERN recommended field-buses. A typical dose
of 50 Gy (5 krad) and 5 * 10**10 n/cm**2 can be obtained in a 6-week
running period.
Bjorn said that the first irradiation period (29/4 to 13/5) had delivered
approximately a gamma-dose of 10 Gy (1 krad) and a n-dose of about 10^10
n/cm^2. These numbers are not yet confirmed as dosimetry is still in
progress.
None of the components was powered during the test.
Bjorn pointed out that the measured radiologic activation of the exposed
equipment had given values below 10 mycro-sivert, allowing for post-irrad.
measurements in the lab.
Most components showed no significant change of el. parameters w.r.t.
pre-rad measurements. In one case the highest "variation" was observed
on
the reference component, which had not been exposed.
Specific findings:
Philips CAN-bus controller interface PCA82C250: no change.
Siemens Optocoupler IL206A: slight change in timing. No significant
change in DC current consumption.
The CTR (current transfer ratio), specified to be within 63% and 125%
has decreased from values around 90% by about 30% absolute for the irradiated
components, but is still inside specs, apart from one unit, which is
54% absolute.- Q from audience about possible anealing: Bjorn will look
into it.
Bipolar bandgap reference AD680JR: spec is 2.5 V with 10 mV accuracy.
Irrad. components have changed by up to 1.3 mV, but are still inside
specs.
The ref. device has only changed by 0.1 mV.
High Accuracy 100 mA low dropout linear reg. volt. ref. ADP3301:
Nominal: 5V. Change: <= 1 mV.
Local Monitor Board LMB Pt100: No significant change.
ATMEL AT90S1200 RISC Processors with flash memory and EEPROM:
Among the 5 devices 3 were OK. 1 device had a changed memory location
in the EEPROM, 1 device was unusable for other reasons.
Discussion: results potentially very interesting. However, need tests
under voltage and reliable dosimetry.- Good candidate for PROSPERO irrad.
test.
4. W. d. Sprolant: First results of magn. field tests on electr. ventilators:
William described the test-setup in the ESS group. A magn. field of
up to 1.5 kG can be generated inside a gap of 200 mm width and 500 mm depth.
This space allows to test equipment like complete power supplies. In
a first round a number of standard electrical ventilators was tested.
Ventilators differ in the arrangement of the magnetic field. In all
tests the rotation speed was tested as a function of the magnetic field
B. All
types showed strong dependance on the angle between rotor-axis and
B. In case of the favourable orientation some types showed nearly no reduction
of
the rotation speed up to 1.5 kG, while all types showed strong reduction
in the unfavourable orientation already around 300 to 500 G.
These results can be understood from the fact that the amount of magn.
material in commercial fans is minimized to just suffice for the anchor
coils. Technical modifications to allow for higher external fields
are conceivable but would certainly increase cost. An attractive alternative
are pressurized-air driven ventilators. These devices are standard
commercial components and should be studied in detail.- William
demonstrated on the basis of the ATLAS and CMS field maps that it is
unrealistic to try to orient all fans in the "favourable" direction.
In the next round of measurements DC-DC converters will be tested. It
was stressed in the discussion that most DC-DC converters need subsequent
filtering in order to reduce noise. Magnetic tests must also be done
for these filters as they often contain inductivities.
==> Next meeting: wednesday, july, 15th, 1998, 10:00h
Robert Richter