HELIX_128_1 irradiation

Gain & Noise:

The gain is determined after measuring the dynamic range. Once this plot was available a line fit was made to find the slope of the input versus output curve. This fit was done in the -10MIP to +10MIP (300u Si) range. The gain is the slope of the fitted line in mV/MIP.

This plot shows the range scan before irradiation:

From this curve one can extract the linearity of the chip. (It has not been corrected for the non-lin. error of the input signal, a "good" signal source is used).
The linearity is calculated like :
data - fitted line
Linearity= dyn. range (for -10 to +10 MIP)

After irradiation :

For each scan a fit has been made (results in prev page). In between the irradiation and measurement steps the bias has not been optimized. So after all steps (160 krad) a new bias optimum was found. This is the red dotted curve in the plot.

7Linearity of the rebiased chip :

Noise:

The noise is determined after capturing 1000 "events". An event is the result of an internal test pulse, ~3.3V. Every event is readout from the same column.
From this data the common mode value of each individual event is subtracted. This really has a high impact on the results and is probably necessary in the experiment in order to obtain low noise (!).

Then the pedestals are determined by subtracting the signal result of the testpulse for each channel. This is averaged over the 1000 events. These pedestal values can also be used for other measurements.

The plot shows all events on top of each other.

Now the standard deviation of the data for each channel is calculated which is a value for the noise.

The next plot is made from the rebiased chip after irradiation.

The main change in bias after irradiation is the change of Vfs. This voltage controls the pulse shape and this has been changed during irradiation. The pulse shape was much shorter after irradiation, and therefor the peak was also shifted.

This resulted in a wrong sampling point of signal at the input of the pipeline. The sampled signal was about 10% lower than the peak value.

If you have comments or suggestions, e-mail me at : r.kluit@nikhef.nl