Direct Injection Hybrid and Extrinsic Silicon

Without preamplification surface state trapping noise will still be a problem for 3-5 pm systems for TDI clocking frequencies over 10 kHz however since the CCD can be fabricated in silicon a buried channel device could be used to circumvent this noise problem. In any case a tapered register is usually used for TDI as shown in Fig. 6.8 to provide a constant fat zero from the background. With this hybrid approach the CCD well width is not directly related to the photoactive area so that saturation problems can be relieved by using a wide register near the end, but this increases the separation requirement between TDI channels, complicating focal plane layout since no more than a 2 1 interlace is desirable. 

The use of a hybrid approach increases flexibility as to detector choice at the expense of packing density and problems associated with inputting the photosignal into the CCD. The photocurrent is first integrated and then periodically dumped into the CCD, which causes a sampling of the signal in the 

The principal difficulty with this input circuit is that since is finite, a fairly 

For currents less than about 4 x 10 amps the MOSFET will be in the weak inversion regime and g =(l/2)eI/kT [6.22,28, 29]. No f problem would exist for 8-12 pm detectors unless the detector capacitance were unusually large as with Pbi Snj(Te diodes. 

The use of an extrinsic photoconductor with direct injection has the advantage that dc gain can enhance I and therefore g, but gain saturation due to sweepout will limit ac gain to 1/2 at frequencies near f [6.30]. The capacitance C of an extrinsic photoconductor can be an order of magnitude lower than for a photodiode, which will lead to a higher /. With an extrinsic photodetector, crosstalk problems must be considered for the detector thicknesses necessary to provide reasonable quantum efficiency, and even for the 3-5 pm window operating temperatures will tend to be below 50 K. 

Even though there are sometimes problems with input to the CCD register all approaches which introduce the TDI function before preamplification have an advantage in that the gain of the TDI register ( /M on the background noise) makes it easier for the post TDI preamplifier to deliver BLIP performance. 

---