Description :
The chips on the ALCT-board were radiated using a 63 MeV proton beam at UC Davis.
For this facility, a current of
50 pA corresponds to a Flux of 0.8532*10**7 protons/cm**2/sec with a Center Fluence
of 0.7850*10**9 protons/cm**2. From this numbers a crossection for the measured SEU (Single
Event Upsets) will be calculated. The plots on the top left show the experimental setup and
on the top right a schematic view of the ALCT-board.
Procedure :
irradiate LCT-Chip0 (Altera) with a current
of 50 pA for a doses of 1 krad.
Got inject-errors (pattern)
and false-trigger-errors at beginning. LCT-Chip shows to be
very sensitive to radiation. After cycling power
all errors disappear.
We have performed an exponential-fit to the plot. Due to the experimental setup there was
always some delay between the start of the irradiation of the chips and the begin of the data-taking.
This gives a dilution of the time-measurements for the first seconds.
Therefore, we didn`t included the first bin in the fit.
From the fit a SEU is expected after 59.2 +- 12.1 rad. This number corresponds to 51.6 +- 10.5 sec using
a dose rate of 1.2 Rad/sec.
It corresponds then to a SEU crossection of (2.3 +- 0.5)*10**-9 cm**2 and a Fluence of
(4.4 +- 0.9)*10**8 cm**2. These results can be compared with the ones obtained from the
OSU group.
Calculation by M. Huhtinen gives for (ME1/1) 6*10**11 neutrons/cm**2
for 10 LHC years (equivalent to 5*10**7 s) at peak luminosity
(see talk)
Throwing in a safty factor of 3 gives 2*10**12 neutrons/cm**2 in 5*10**7 sec of peak LHC lumi
(see talk)
went up to total dose of 2 krad.
Showed same effect.
testing of flash-ram 1 in three steps :
dose of 4 krad with a beam current of 200 pA.
dose of 4 krad with a beam current of 200 pA.
dose of 8 krad with a beam current of 400 pA.
After each step verification information of flash-ram
showed no errors. Flash-ram showed to be radiation
resistant up to 16 krad.
testing of LCT-Chip 2 in three steps:
dose of 4 krad with 400 pA beam current.
dose of 4 krad with 400 pA beam current.
dose of 8 krad with 800 pA beam current.
Got same errors as before (inject and false trigger errors)
at the beginning.
irradiate flash-ram 3 with beam current of 2 nA and up
to a total dose of 15 krad.
Verification of information
in flash-ram didn`t show any errors.
irradiate DACs and ADCs with a beam current of 2 nA up
to a total dose of 15 krad. Temperature went down
to 33.4 and stays there.
DACs
and ADCs damaged.
irradiate slow-control chips with a beam current of
2 nA up to 15 krad.
inject- and false-trigger-errors.
After cycling power everything working again except for
the temperature sensor.
Next measurement : Look for latch-ups.
continue irradiating slow-control chip.
Increased current to 5 nA and total dose to 100 krad.
After $\approx 60 $ krad lost connection to board. No JTAG-chain
anymore.
No latch-up
irradiate LCT+flash-ram for a total dose of 100 krad.
No latch-up. The 4.0 V current went over 3 A.
(normally: 1.9 A).
irradiate ADCs and DACs for a total dose of 100 krad.
No latch-up. The 5.5 V current went up to 0.6 A.
(normally: 0.3 A).
continue irradiating in steps of 100 krad up to 300 krad.
No latch-up. The 5.5 V current went up to 0.666 A
and stays there.
Afterwards same procedure for LCT+flash-ram. The 4.0 V current
went up. No latch-up.
Next measurement : Estimate power-cycling
LCT-chips showed to be very sensitive to radiation. After errors occur it is
necessary to cycle the power. To determine the time after which the power needs
to be power-cycled at LHC, new tests were necessary. In this test the beam was collimated
to one of the four ALCT-boards. A beam-current of 50 pA was selected. The LCT-chip got then
radiated until an error occured. Then the irradiation was stopped and the power of the board was
power-cycled. This procedure was repeated 100 times.
The plot below shows the distribution for the radiation-dose necessary to get
an error in the LCT-chip.
