CW_irradiation

Download Report

Transcript CW_irradiation

Irradiation test of CW bases
Yu. Gilitsky / IHEP (Protvino)
1/10
The irradiation test was done to compare performance of different CW
configurations at high doses.
The test was conducted 12-18 Nov 2008 at the HiLum facility (IHEP, Protvino)
constructed for the irradiation tests of the ATLAS LAr.
The 50 GeV proton beam (extracted using bent crystal technique) has intensity
up to 1012 p/spill (9 sec interval, 1.7 sec duration)
Four samples with different configurations were places behind Absorber I (1λI
steel). UCTRL was fixed to ~2V by a resistor divider arranged on each board. The
output HV was monitored during the whole irradiation period.
ECAL CW samples
2/10
Proton beam 50 GeV
Actual beam position
Expected beam position
Absorber I (1 λ0 of steel)
D1
CW1
D2
CW2
D3
CW3
D4
CW4
D1-D4 – film dosimeters
CW #
config
details
krad
CW1
HCALmc2
Cf=100pF, Rf=62kΩ, CJ=47pF, C1=C2=0, RJ=0
800
CW2
ECAL orig
C1=C2=47nF (C1MV), Rf=0, Cf=0, CJ=0, RJ=0
900
CW3
HCALmc1
Cf=100pF, Rf=62kΩ, CJ=47pF, C1=C2=0, RJ=5.6kΩ
700
CW4
ECALmc
C1=C2=47nF, Rf=62kΩ, Cf=100pF, CJ=0, RJ=0
500
3/10
Protons rate from beam monitors
start of degradation
CW1 and CW3
start of irradiation
4/10
HV, V
Significant HV drop was observed in
both HCALmc bases after ~500 krad;
the ECALxx bases were not affected.
In order to proof that the drop was
determined by the configuration, in 5 days
after the irradiation the CW2 base (900 krad)
was modified to:
▪“ECAL+HCAL” (ECALorig +
Rf=62kΩ+Cf=100pF): OK
900
800
700
600
500
400
CW1
CW2
CW3
CW4
300
200
100
0
12
13
14
15
▪ECALmc (C1=C2=47nF): OK
16 17 18
November 2008
▪HCALmc (C1=C2=0): HV≈670V, unstable
▪ECALmc: OK
High rate test (with PM) in the ECALmc mode:
▪ HV=1500V stability checked up to 30 μA OK
▪ HV=850V – checked up to 50 μA OK
One can conclude that configurations where C1
is present (connected either to MV or to
ground) (“ECAL” type) are more “rad hard”
than those where C1 is absent (“HCAL” type)
Summary on HVout in “HCAL” modes
HVout in HCAL mode, V
base
expected
right after
irradiation
in 2
days
CW1
888
345
550
CW2
886
CW3
837
CW4
835
in 5
days
in 50
days
730
670
410
544
720
5/10
On 08/01/09 (~50 days after the irradiation) the CW1 base (800 krad) was
studied in details in both “HCAL” and “ECAL” configurations. Switching was done
by soldering / unsoldering capacitors:
“HCAL”: CJ=47pF present, C1 and C2 absent
“ECAL”: CJ absent, C1=47 nF and C2=3.3nF present
always present were: RJ=10kΩ, Rf=62kΩ and Cf=100pF
HVOUT as a function of UCTRL
Conditions: MV=80V, LV=+6V -6V
Same conversion factor, but different
max HV, in “HCAL” and “ECAL” modes.
For more detailed studies 4 sets of
oscillograms were taken:
# conf UCTRL, V HV, V
1
2
3
4
HCAL
HCAL
ECAL
ECAL
1.174 502
≈4.4 730
1.756 783
2.495 1140
*saturated*
Each set consisted of ~20 oscillograms taken
in different points, such that it was possible
to calculate currents in places of interest.
6/10
set #2
set #2
set #4
set #2
As a result of irradiation:
▪The oscillator produces 25 kHz sine-like wave, 4 V p-p.
▪op amp gain is ~2-5 at 25 kHz (UOUT(t) and ΔUIN(t) of
the oscillator op amp shown; note gain = UOUT/ΔUIN ≈ 5
and ~90o phase shift of UOUT wrt UIN)
▪ DC gain is at least several hundred: < 20 mV change in
ΔUIN flips the output (measured later 14/01/09)
set #2
▪op amp gets significant output impedance (output current
limitation?); differential output impedance of saturated
regulator op amp shown, set #2
▪npn transistor β’s become 5-7 (Q1 & Q2 shown)
7/10
Looking at “HCAL” set #2, to find out the reason for limiting HV
IPUMP (current in RJ=10k):
Q3 part (up) OK, problem
with Q2 part (down)
Q1 is always saturated, Q2 active
 insufficient current into Q2 base
Q1 IC
NB
IC negative
(reverse)
for part of
period
8/10
Compare to “ECAL”
set #4,
IPUMP: both Q3 and Q2
half-periods OK
HV=1140 V
Factor of >2 wrt “HCAL” in charge
injected into the Q2 base
9/10
Possible explanation of the difference between “HCAL” and “ECAL” type bases
Because of low gain of irradiated transistors, the irradiated oscillator op amp (with
lowered load capabilities) has to draw non-negligible current into Q2 and Q3 bases.
The presence of capacitor C2b means that the amount of charge delivered into
[Q2 b-e junction || R2e] during the half period when Q2 is open should be returned
back during next half period, when it is closed (Q3 open)
In the “ECAL” configuration, C2b is recharged from C1 through R2e and R2b. In the
“HCAL” configuration, the return path is more difficult: as C1 is not present, charge
comes from the regulator op amp, passing through R1b, b-c junction of Q1 and then
R2e and R2b (this is the reason why in the “HCAL” configuration the Q1 collector
current is reversed during ~ half a period). And max possible returned charge is
insufficient for operation at higher HV.
As a proof of correct understanding the problem: to boost
the C2b recharge in the “HCAL” configuration, a diode was
added || R2e. The max HV increased till the same 1320 V as
in the “ECAL” configuration.
10/10