Radiation Testing of
TMP36 Temperature Sensors in a Proton Beam
E. Hazen - 15 April, 2003
PRELIMINARY
Test Setup
A test board (see photo) was designed which supports (10) TMP36 devices in SO-8
packages. A 5m ribbon cable connected this board to a 32-channel, 12-bit ADC, which
digitized the 10 TMP36 outputs, along with the 3.3V power supply voltage.
Irradiation Procedure
The test board was placed in a 157.6 MeV proton beam at the Northeast Proton Therapy
Center (at Mass General Hospital in Boston). The beam size was about 2.5 cm which
was large enough to provide uniform irradiation of all devices to within about 10%. The
radiation was delivered in 15 “runs”. The first 5 runs delivered 10krad each over a
period of about 1 minute per run. The ADC values were recorded every 5 seconds. The
remaining dose was delivered in 10 additional runs of 25krad each. The total elapsed
time was about 30 minutes (including down time between runs).
After the test data was recored for about 5 minutes, then the irradiated board was
disconnected. The board was then stored at room temperature for 6 weeks.
Test Results
During irradiation, all 10 device outputs were recorded at 5 second intervals, with data up
to 30 kRad shown in Plot 1 below. Unfortunately, the ambient temperature in the
radiation area was not independently monitored, but before the test it was 20.8 C and
after the test it was 21.4 deg C (average reading of 10 non-irradiated sensors). Plot 2
shows the data averaged by run up to the full dose of 300 kRad.
TMP36 During Irradiation
33
32
31
30
T0
T1
T2
T3
T4
T5
T6
T7
T8
T9
Deg C
29
28
27
26
25
24
23
22
21
20
0
10000
20000
30000
TID rads
Plot 1. TMP36 Outputs During Proton Beam Irradiation
The 8 curves (T0...T9) represent the 10 sensors. The gap in the values around 10krad
represents the transition between the first and second cyclotron runs.
TMP36 Output vs TID
80
75
70
T0
65
T1
60
T2
T3
55
T4
T5
50
Deg C
45
40
T6
T7
35
T8
T9
30
25
20
0
100
200
300
kRad
Plot 2. TMP36 Outputs During Irradiation (averaged by run)
TMP36 Calib. after 6 wk Room Temp Anneal
75
70
R2
T0
T1
T2
T3
T4
T5
T6
T7
T8
T9
Sensor Deg C
65
60
55
50
45
40
35
30
25
20
20
30
40
50
60
Ref Deg C
Plot 3. Calibration After Room Temperature Annealing for 6 Weeks.
After irradiation, the boards were stored at room temperature for 6 weeks. They were
then calibrated in a temperature-controlled box. The results are shown in Plot 2 below.
The absolute calibration of the irradiated sensors is drastically shifted, but the relative
calibration is still fairly good. Plot 3 and 4 show the calibration after 168 hour 40 deg. C
annealing and 12 hour 100 deg. C annealing, respectively. Essentially no recovery is
observed.
TMP36 after 168h 40 Deg C Bake
70
65
Sensor Deg C
60
R2
T0
T1
55
50
T2
T3
45
T4
T5
T6
40
35
T7
T8
30
T9
25
20
20
30
40
50
Ref Deg C
Plot 4. Calibration After 168 Hour Annealing at 40 Deg C.
TMP36 After 12 hr 100 C Bake
70
65
Sensor Deg C
60
R2
T0
55
T1
T2
T3
50
45
T4
T5
40
T6
T7
T8
35
30
T9
25
20
20
30
40
50
Ref Deg C
Plot 5. Calibration after 12 Hour 100 Deg. C Annealing
Conclusions
The TMP36 temperature sensor appears to lose absolute calibration after about 10krad
equivalent dose in a 167MeV proton beam. Immediately after delivery of a 300 kRad
dose, the readings from the 10 test devices were widely scattered from 35 to 80 degrees
C. After room-temperature annealing, the readings are much more tightly clustered with
a total spread of about 5 degrees C. Subsequent high-temperature annealing had no
significant affect.