Wednesday, November 02, 2011

e2v Claims Record QE in 5.3um FSI Pixels

e2v begins sampling in this quarter the ‘Ruby’ family of image sensors (EV76C660 and EV76C661), which is said to have sensitivity performance beyond that considered possible on traditional front side illuminated sensors, with a pixel size of 5.3µm. The reported QE is over 80% in visible spectrum and over 50% in NIR at 850nm. The sensors are targeting industrial and machine vision applications, as well as very low-light imaging in outdoor camera applications.

Update: A separate PR on Ruby sensors has been released.

13 comments:

  1. is this sensor made by towers?

    ReplyDelete
  2. No, i believe it's made by sky scrapers.

    ReplyDelete
  3. Not sky scrapers, but rather a stand of cypress trees.

    ReplyDelete
  4. Actually its made by light pipes

    ReplyDelete
  5. 5.6um state-of-the-art pixel reach that level of performance (with microlens). Not impressive.

    ReplyDelete
  6. what is the state of the art QE in a 5.6um pixel??

    ReplyDelete
  7. Hey guys, now that we do get some numbers, please do not complain too much about it ! A QE of 80 % in the visible is not that bad at all. I am pretty sure that my mobile phone camera as well as my digital SLR sensor do not show such a high QE. Question is whether this 80 % is the QE for the complete pixel or just for the diode ?

    ReplyDelete
  8. LOL at the comments on sky scrapers and towers! made my day.

    ReplyDelete
  9. Probably high NIR QE (>50%) = bad crosstalk as well. Internal QE@850nm at 50% requires >13 um epi with an optimal anti-reflective coating..

    ReplyDelete
  10. There are smaller (<4um) pixel products that have real QE (% of photons captured by pixel) greater than 80%. I know at least one major image company that samples.

    ReplyDelete
  11. @ "Probably high NIR QE (>50%) = bad crosstalk as well."

    Can be, but not necessarily. One can make lightly doped thick epi and apply high negative voltage to the backside. Thus depletion can extend to a large depth and crosstalk would remain in-check.

    ReplyDelete
  12. Vladimir,

    agreed, but i'd hate to be the one to measure the dark current...

    or at least the the that had to report it to management/marketing folks! :)

    ReplyDelete
  13. Well, I'm using the EV76C661 and it does have much higher dark current than previous "sapphire" line. I've seen a cam test of the *60 and it also looks very poor for (planetary) imaging in IR at high gain an bit longer exposures). Such high QE, but a lot of noise. Strong peltier cooling could probably help, but I don't know if that would limit the noise as much as needed.

    ReplyDelete

All comments are moderated to avoid spam.