Each strip or pixel of a silicon sensor must be connected to its own readout channel. This can be done in two different ways. Either, there is a direct connection from the strip or pixel to the amplifier input (DC coupling). This implies that the amplifier must sink a fraction of the detector leakage current that corresponds to the strip or pixel. While this contribution is usually small on pixels because of their limited size, it can be much higher than the signal current especially with irradiated strip detectors. Since the leakage current depends on the applied bias voltage and the radiation damage, it is difficult to build an amplifier which can handle such a wide range of input current. The solution is to bypass the DC leakage current over a resistor and pick up only the AC part over a capacitor (AC coupling). Obviously, this is the preferred technique for present silicon detectors.
With silicon strip detectors, resistors and coupling capacitors are usually integrated
into the sensor. The bias resistor is either implemented as a field oxide field
effect transistor (FOXFET) or, more common, a polysilicon meander structure, which is
less vulnerable to radiation damage. The capacitor is built by a metal layer over silicon oxide
on top of the strip implant.
The bias voltage, which is applied to the backplane, is usually decoupled with an RC filter.
Fig. shows the connection schemes for a single channel of a silicon detector
with both DC and AC coupling methods.
![]() |