Wednesday, December 24, 2014

Alexima Applies for GS Pixel with Tri-State Transfer Gate Patent

Alexima's patent application US20140367552 "Image sensors, methods, and pixels with tri-level biased transfer gates" by Jaroslav Hynecek and Alexander Krymski proposes a global shutter "pixel with a transfer gate that is controllable among at least three biasing conditions, including a first biasing condition in which electrons are transferable from a photodiode to a potential well under the transfer gate, a second biasing condition in which the electrons are confined in the potential well under the transfer gate, and a third biasing condition in which the electrons are transferable out of the potential well under the transfer gate." The pixel cross-section and its band diagram in all 3 modes is shown below:

9 comments:

  1. Nice idea, the combination of a 2-phase CCD gate (under the transfer gate, also present in the IMX174 of Sony) and a virtual gate CCD (region 207). Not surprising because the inventor of the original virtual phase CCD is also mentioned as the inventor of this patent application, being Jerry Hynecek.

    ReplyDelete
  2. What is the utility of 205 please??

    ReplyDelete
    Replies
    1. It's just a way to create different dosage, so that PD area gets 511 and 205 dosage together, while parts of AB and Tx gates get only 511.

      Delete
  3. I also like this idea but there may be a drawback or two.

    It is similar in structure to a previous Aptina paper co-authored by Jerry and presented at the 2013 IISW. http://www.imagesensors.org/Past%20Workshops/2013%20Workshop/2013%20Papers/Slides/12-1_Velichko_SLIDES.pdf
    (see slide #19).

    This 2013 IISW paper also inspired the "pump-gate jot" that we proposed in our 2014 IEDM paper as a concept and which is in fab.

    Actually, I hope the Velichko et al paper wins the WKA. It was very good, in my opinion at least!

    ReplyDelete
    Replies
    1. Indeed, the TX shown here is effectively a "pump gate" operation except that the well depth difference is created by the P doping and in Aptina case, it is created by the N doping on the opposite side.

      -yang ni

      Delete
  4. Sony, Aptina, Alexima.. others? How far is it from real product in Aptina and Alexima case?

    ReplyDelete
  5. I remember well that in SONY interline CCD, one of the vertical phases use 3-level signal. One (low) for barrier during integration, one (high) for charge readout and other (mid) for vertical charge shift.

    -yang ni

    ReplyDelete
  6. Oh, duh. This structure would be very good for ToF sensors with CDS output, especially with multiple output ports tied to the same PPD (like our old Photobit patent Alex!).

    ReplyDelete
  7. Middle level does not pin the surface and dark current will be huge while containing charges.

    ReplyDelete

All comments are moderated to avoid spam.