Invisage launches an updated web site featuring a new Technology page:
Invisage sensor cross-section. Note, there is no microlens.
The technology page also introduces Dynamic Zoom feature - a sort of binning in continuous photoconductive film. Nice idea. I hope they find a way to combine it with CFA:
I am not sure why given enerything similar, the QE in blue is less for silicon that the photoconductive film? The QE graph looks as if there is a simple offset; I would expect the QE curves to overlap in the blue and part of green (limited by surface recombination) and separate in red (limited bulk thickness) .Probably this is a simple simulation, without taking into account the surface recombination effect (both front and back-surface)
I don't know why Quantum film still need the CFA on the top. Originally, I thought, Quantum film itself has the detectability over specific color range ? In this case, this Quantum film detectable range is around whole visual wavelength range with even high QE. Hmm... it just can replace Silicon so, coating process might be more merit compared to Silicon Photodiode process. In this coating process with Quantum film, how they can be sealed to avoid outgassing of this film ?
I do wish Invisage a lot of succes in solving the dark issues with this structure. Think about amorphous silicon on top of CCDs (Toshiba, 90's), about amorphous silicon on top of CIS (Silicon Vision, 90's) and about germanium on top of CIS (Noble Peaks, 00's).
The dark issues have nothing to do with the bandgap of Ge, but the problem is lying in the contact to the silicon. This is killing your dark performance. BTW, that was also the case with the classical 3T APS device. Because there is NO electrical or physical contact to the PPD in a 4T pixel, it has such a great dark performance !
Maybe if no CFA the Dynamic Zoom could in principle be used to provide image stabilization without moving parts?
ReplyDeleteI am not sure why given enerything similar, the QE in blue is less for silicon that the photoconductive film? The QE graph looks as if there is a simple offset; I would expect the QE curves to overlap in the blue and part of green (limited by surface recombination) and separate in red (limited bulk thickness) .Probably this is a simple simulation, without taking into account the surface recombination effect (both front and back-surface)
ReplyDeleteI don't know why Quantum film still need the CFA on the top. Originally, I thought, Quantum film itself has the detectability over specific color range ? In this case, this Quantum film detectable range is around whole visual wavelength range with even high QE. Hmm... it just can replace Silicon so, coating process might be more merit compared to Silicon Photodiode process. In this coating process with Quantum film, how they can be sealed to avoid outgassing of this film ?
ReplyDeleteThis doesn't worth any attention, totally meaningless.
ReplyDeleteI do wish Invisage a lot of succes in solving the dark issues with this structure. Think about amorphous silicon on top of CCDs (Toshiba, 90's), about amorphous silicon on top of CIS (Silicon Vision, 90's) and about germanium on top of CIS (Noble Peaks, 00's).
ReplyDeleteRegarding Noble Peak's Ge process, the dark current was lower than bulk Ge. Cooling was required due to the small bandgap of Ge.
DeleteThe dark issues have nothing to do with the bandgap of Ge, but the problem is lying in the contact to the silicon. This is killing your dark performance. BTW, that was also the case with the classical 3T APS device. Because there is NO electrical or physical contact to the PPD in a 4T pixel, it has such a great dark performance !
Delete