The motivation for the TCC2 tests see the seminar presentation of Raymond Rausch, general.ppt (power point 49kb). The radiation tests at the TCC2 area started in 1998. However in 1998, the electronics were not powered during the tests. Despite this some interesting results were obtain especially concerning the optocouplers and EEPROMs, see radtst3.html
1. LMB Measurement program for 1999.
For 1999 tests an 'On-Line Radiation Test Facility'
has been installed with 150 m of cables from the underground area TCC2
area to the counting room in 889 R-011. All equipment is continuously
powered and can be readout with the help of the CAN bus. The ATLAS DCS
setup consists of three CAN nodes installed in the radiation area. They
are powered with two +15V power supplies placed in a rack behind shields
at about 20 m distance. From this rack 150 m of cable connects the CAN
bus to a rack in the room 889 R-011. From there to the computer CAN interface
there is a10 m cable. The CAN bus has a total length of about 180
m. The readout software on the PC is the National Instruments BridgeVIEW.
The LMB
application software has been written by S. Filimonov, R. Mattsson.
The status of the experiment can be monitored (at CERN) via web-access
as shown in this display!
The CAN nodes that are installed at the on-line test facility has the following functions:
1) test of a standard 16 channel LMB,
2) test of two types of optocouplers,
3) test of microcontroller for single effect events.
1.1 Standard 16 channel LMB (16ch LMB)
This node consists of a CAN module and a 8 channel Pt100 module. The
modules contain the same components as during the 1998 test. However, based
on the results from 1998, the circuit gain of the optocouplers have
been increased by a factor 8 by changing the collector resistors at the
output of the optocouplers a factor 6, see
schematic of ADC module With these changes it is expected that
the LMB will stand about 10 12 neutrons cm-2
- limited only by the optocoupler. (The tests performed with
neutrons at PROSPERO show that all components except the optocouplers in
the CAN node and ADC works up to at least 9 10 12 neutrons cm-2
). The 8 inputs are used to monitor the following sensors: 3 high
precision 100 ohms resistors (to check the stability of the LMB), 3 temperature
probes measuring the temperature at the LMB environment and 2 voltage monitoring
of the analog and digital LMB power supplies.
1.2 32ch Test of optocouplers (32ch LMB)
This node consists of a CAN module, a 32 channel LMB ADC module and
two adapter modules, see Fig.1. and Fig.
2. These are equipped with 4 dual optocouplers each. The ADC module
is modified and without optocouplers and is therefore capable of withstanding
radiation effects (se above). The optocouplers chosen
for the tests this year are MOCD223 and
HCPL-0731 . Both types have high gain and can
be used with low input diode currents.
1.3 AVR Test of microcontrollers (AVR)
In order to test for errors caused by radiation effects in the memories
of the microcontrollers used in the LMB the following tests are performed.
6 microcontrollers of the type AT90S2313,( 2k flash memory and 128 bytes
of EEPROM memory) have been mounted in special box, which can be read out
by a standard LMB. A unique test pattern has programmed in each of the
AT90S2313. The microcontrollers are put in its idle state when it is not
read out by a few lines of code in first addresses of the flash memory.
The program in the LMB checks for differences from a predefined test pattern
in each word read out. In its initial state there 14 words per microcontroller
which differs from the test pattern.
The equipment above was tested during three periods of time. During the first period the readout stopped working after 3 days of full intensity beam, which was found to be due to a power supply failure (not under test). The power supply was mounted in the radiation area in what turn to be the most irradiated area. The following results were obtained:
2.1 Optocouplers [32 ch
LMB and 16ch LMB] (three different types
all are dual devices SMD SOIC-8)
|
|
|
|
Comments |
|
|
|
|
see the online measurements: optores.htm |
|
|
|
|
-"- |
|
|
|
|
used in the LMB, measured after the test was interrupted |
The total dose is estimated (from the radiation monitors and the conversion factors supplied by TIS, see radmon) to be 300 Gy and 3 1012 n/cm2 equiv 1Mev Si.. The dose rate as function of time is shown in optotime.html
Comments: The 16 ch LMB was working
until the CTR of the optocouplers (ILD206) was reduced to about 3.5% of
the original value. The HCPL type of optocoupler is significantly better
that the other types.
2.2 Multiplexer ADG439FBR [32 ch LMB and 16ch LMB]
These devices are used in 32ch LMB and 16ch LMB. ( SMD SOIC-16) Fault protected analog multiplexer (35V) .
Test 1) The switching between channels stopped
and finally no channel worked.
Test 2) The devices were replaced with the
same devices that was tested 1998
but failed again after a few days of operation.
|
|
Devices type | No. of units measured | Error observed | Comments |
|
|
|
ADG439FBR |
|
Channel selection | Date code is 9843,9845,9845 |
|
|
|
ADG439FBR |
|
Channel selection | Date code is 9604 |
|
The node stopped working due to increase power consumption (16
times). The chips were
still working except that it was not possible to erase
the contents of the flash memory.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The online monitor system detected the following events.
|
|
|
| (3 occasions in the registers the CAN controller SAE81C91) | The CAN node stopped operating. It did not reply on CAN messages. They could be put into normal operation by turning the power OFF and ON. (This is equivalent to hardware rebooting of the CAN node). |
| EEPROM or flash (number of devices =12) | No erasure of the flash or EEPROM memories was observed. |
