A barrel kapton hybrid populated with 12 ABCD3T chips was irradiated with
neutrons in the reactor in Ljubljana.

SETUP

For this irradiation we use Mustard-Slog readout with SCTDAQ software.
Other components are same as before: home made bench power supply,
Melbourne LVDS repeater card, long cables (5m between the hybrid and
LVDS repeater, 3m between the repeater and VME crate).

The main difference compared to our previous irradiations is that now we
cool the hybrid during irradiation and also monitor the temperature on the
hybrid with thermistors. We have water cooling system with the cooling block
attached to the aluminium plate on which the hybrid is mounted.
The temperature on the hybrid during irradiation (and measurements) was
between 25 and 30 C.

RESULTS

Due to the problems with cooling we actually did 8 irradiations instead of 4.
In the first 4 irradiations the hybrid didn't enter the core properly so we only
reached about 2.5*10^13 n/cm2 with certain fluence gradient across the hybrid.
Then in the last 4 irradiations we did the job properly and reached the total
fluence of 2*10^14 n/cm2. The first four irradiations are the reason that the
fluence is not the same for all the chips (but the differences are small at
higher fluences), as one can see in the plots.
Also one can see on the plots below that there is a kink in the slopes of gain an noise
vs. fluence curves between the first and second irradiation period.
This indicates that the dosimetry of the first irradiation might be wrong,
although we don't know what could be the reason for this. The only way to
find out if this was the case is to irradiate another hybrid.

Trim DACs:

The trim DAC curves for all ranges for all chips before irradiation  and
after full irradiation It seems that chip number 8 had problems with setting
the correct trim DAC range after irradiation.
The distribution of 50% points after trimming is shown in this plot.

Response curve:

Response curve with 10 charges: 0.5, 0.75, 1, 1.25, 1.5, 1.75, 2, 3, 4, 6, and 8 fC is
shown here.

Time walk:

Time walk before and after irradiation. The curves after irradiation are shifted up
so that they fit into the same plot as those taken before irradiation.

Gain and Noise:

Gain and noise was masured from threshold scans done with 4 charges , 1.5, 2 and 2.5 fC
at shaper currents 20 and 30uA and preamp currents 100, 150, 200 and 250uA.

Gain and noise across the hybrid for bias currents:
Ipreamp = 150uA , Ishaper = 20uA ("nominal case")
and Ipreamp = 100uA , Ishaper = 30uA  (best case).

Here are the gain and noies evolution with fluence for the two
biasing cases: Gain(150,20), Gain(100,30),
                      Noise(150,20), Noise(100,30)
 

At preamp currents higher than 150uA gain gets very low after irradiation.
The response curves after full irradiation for charges 1, 1.5, 2 and 2.5 fC for all
the chips at different biasing conditions can be seen here.

Preamp current and Vcc:

At higher preamplifier currents the chips are not working very well after irradiation
because gain drops to virtually zero and also threshold spread gets very high.
One of the reasons for this seems to be the value of Vcc.
When doing scans at different preamp and shaper values, Vcc is not adjusted at
the power supply. Thus Vcc is lower at higher values of preamp and shaper currents.
The influence of the value of the value of Vcc is shown in the plots of
means and gains for link 0 after irradiation. It looks that higher Vcc doesn't have much
effect for 150uA of preamp current whereas at Ipreamp = 200uA the influence is more
obvious.

Power consumption:

Voltages were Vdd=4  and Vcc=3.5V (measured when currents are as below)
Currents:
Idd = 465mA at Vthreshold = 500mV (did not change with fluence)
Icc = 917mA at Ipre=220, Ish=20 before irradiation and
         730mA after 2*10^14n/cm2.
 

CONCLUSION:

It seems that the chips in this irradiation suffered more compared to the
irradiation at the PS (UKK2 - PS Oct00) .  One explanation for this would
be that the dosimetry of the first irradiation period was wrong for some
unknown reason as mentioned before. Although we did measure the fluence
with  dosimetry diodes and we don't know for a reason for not thrusting those
measurements the kink in the gain (noise) vs fluence curves indicates
that the actual fluence might be higher for the first period.
Since we can thrust the dosimetry of the second irradiation period one can
assume that the slopes of gain and noise vs. fluence curves for this part are correct.
With those slopes over the whole fluence interval the gain
at 2*10^14 n/cm2  would be ~30mV/fC and noise ~900electrons which is
approximately in agreement with PS irradiation.
But to confirm this hypothesis we would need to irradiate another hybrid.

After irradiation with neutrons we irradiated the hybrid also with gammas
from the reactor. Click here for the short report about this irradiation.