News and discussions about image sensors
Albert, how many pixels are lost normally when 2 sensors are butted together please ? Thanks !!-yang ni
It all depends on the pixel pitch of the sensors. But in many cases the loss in pixels can be limited to 1 single pixel.
I've seen a very large linear array, which was butted. The pixels next to the cut edge were of different size - probably to compensate for loss of sensitivity. Pixels pitch was significantly larger, than in current imagers, so the loss would be of "sub-pixel size"
Sometimes a SMALLER pixel is put at the butting edge. In this way, at least you get a bit of sensitivity in the edge pixels. With a gain correction for the edge pixels, the loss can be minimized to "sub-pixel", but in that case we are talking about pixels of at least 25 um or larger.
Albert, what are the special dicing techniques employed to keep the pixel loss to a single pixel? How is chipping damage avoided? How does one obviate the need for a guard ring?or are you talking about 100+um pixel dimensions?
Sometimes special/extra precautions are taken during the dicing. Or, even a combination of a wider trench etch + dicing is used to prefectly define the edge of the die and to avoid chipping. I am not talking of 100+ um pixels, but even with pixels of 25 um, a loss of maximum 1 pixel is possible.
Albert do you talk about plasma singulation method pleas? Is this popular in dicing industry pleas? thank!!
Since Olympus published a paper on its 2MP sensor that has GS w FPS=10000, I have wondered how Olympus could "stich" the 10 of these to get 20MP sensor w similar FPS and GS. Glad to see others are tackling a more general problem.
Thanks for your comments and your support. More to come about various stitching technologies after the holidays.
All comments are moderated to avoid spam.