Monday, March 09, 2020

Omnivision Nyxel-2 Technology Extends NIR QE to 50% @ 940nm and 70% @ 850nm

PRNewswire: OmniVision announced Nyxel 2 — the second generation of its NIR technology. Despite launching the first generation more than two years ago, competing mass-produced CMOS image sensors are said to be still failing to achieve comparable NIR performance. Meanwhile, OmniVision’s R&D team has continued to refine its pixel architectures and processes to achieve new records in QE, with Nyxel 2 now providing a 25% improvement in the 940nm NIR light spectrum and a 17% bump at 850nm wavelength.

Nyxel 2 technology further extends OmniVision’s leadership in NIR image sensing,” said Lindsay Grant, SVP of process engineering at OmniVision. “Pushing the envelope of NIR performance opens new possibilities for applications that operate in near or total darkness, including more accurate driver-state monitoring, better surveillance capabilities for security systems and new under-display sensing applications for mobile devices.

Competing CMOS approaches for NIR image sensing continue to rely solely on thick silicon to improve NIR sensitivity. However, this results in cross-talk and impacts MTF. Attempts to overcome this by introducing DTI often lead to defects that corrupt the dark area of the image.

With Nyxel 2, OmniVision has further refined its approach to NIR imaging that combines thick-silicon pixel architectures with careful management of wafer surface texture to improve QE, along with extended DTI to retain the MTF levels of the first generation without affecting the sensor’s dark current.

With these refinements, OmniVision’s Nyxel 2 can now achieve 50% QE at 940nm — a 25% improvement over the first generation, as measured using data from a 2.9um pixel. At the 850nm NIR wavelength, Nyxel 2 can provide 70% QE, which is not only a 17% improvement over the first generation, but it is now on par with the QE levels of top RGB sensors that operate with visible light.

OmniVision’s first image sensors with Nyxel 2 technology are expected to be available during the second half of 2020.

For comparison, the first generation Nyxel announced in 2017 had a QE below:


  1. Headline should read "70% @ 940nm and 50% @ 850nm"

  2. No, not regarding the first figure

  3. The drawings imply that to get these QE numbers, the incoming light must be collimated. If so, the numbers are meaningless with most optics unless a microlens lens falloff curve is also supplied.

  4. I think the drawing implies that there are light scattering structure under the Microlens combine with the DTI to "increase" the distance of silicon the light has to go through.

  5. Can someone explain why a BSI sensor having Lens Array. Since the whole pixel area serves as sensitive area. Also there's no obscuration of metal other components on the surface.

    1. Have been wondering that too ! :)

    2. DTI and other stuff to reduce optical crosstalk take optical area. ulens needed to compensate for such loss


All comments are moderated to avoid spam and personal attacks.