Six n-type ABCD2 were irradiated bare i.e. before they were mounted on hybrid:

• 2 chips to 1x10¹³ n/cm² .
• 2 chips to 6x10¹³ n/cm² .
• 2 chips to 2x10¹⁴ n/cm² .

After irradiation one chip for each fluence was mounted on the Freiburg kapton hybrid used for irradiation of biased (and clocked) ABCD2 chips.

Performance of chips irradiatied biased was compared to the performance of chips irradiated unbiased. On the plot gain as a function of preamplifier bias for the two biased chips can be compared to the gain of unbiased chips irradiated to the same fluence.

Conclusions:

Chips irradiated unbiased show similar fluence dependent behaviour as chips irradiated biased. For both the range of preamplifier settings at which chips work shifts to lower values as the fluence is increased. Although there are some differences - the unbiased chips seem to be damaged more (at 1x10¹³ n/cm² and 6x10¹³ n/cm² ) than biased ones - which need further studies the irradiation of unbiased chips seems to be a promissing method for irradiation of a large number of chips.

Thus, in case there is going to be a massive chip irradiation programe, e.g. for QA, the irradiation of unbiased chips could be the method because it is much simpler to do irradiation of bare chips compared to the irradiation of chips mounted on hybrids. This is mostly because the radioactivity of the hybrid and support structures after irradiation is a serious problem which would make irradiation of a few hundred chips a very long and difficult task.