I think I’m finally getting the hang of this blogging thing – and while we have a requirement not to prattle on about IP, I’m pretty sure what I’m about to share with you will be announced shortly anyways.
We’re in the design phase of a revolutionary, next generation image sensor to address a multitude of markets and benefit machine vision system designers from the factory floor to large scale OEMs. Expectations are that this latest chip will fully eclipse today’s technology for general machine vision – especially for food and environmental inspection. The chip will be available in 3 resolutions, 1, 4 and 12 Megapixels with two primary pixel sizes – 5 and 25 µm. Starting with the base pixel, the sensor field-of-view can be adjusted to meet most requirements, covering a range of 5 -100µm pixel pitch.
As a result of the properties of the silicone, pixel size can be finely tuned by camera or module design, and based on the theoretical properties of the underlying technology we estimate pixel size could vary from 1 to 10x base. In addition, with the current state of the art we are expecting to demonstrate a 1-4x expansion ratio.
“I am over the moon about this latest achievement,” commented Teledyne DALSA’s Vice-President of Imaging. “This proves once again that Research and Engineering can work side-by-side in the development of market-leading innovation. It won’t be a stretch to convince customers that their static optical systems are history. Our next big challenge will be to motivate lens manufacturers to build real-time adaptable Field-Of-View systems to make full use of our dynamic flex devices.”
Working with unique materials brings technological challenges that need to be addressed. For one, pixel fill factor takes a hit, as we need to increase the metal density for the backend process stack. As we optimize for expandability the metal volume has to be increased to allow for a wide range of aspect ratios. If not accounted for properly, open connections and sensor failure can result. To improve device lifetime with application cycling, a new process line will add aluminume to the backend. Its electromechanical properties lend themselves perfectly to the architecture of the new flex chip.
The current device suffers lifetime reliability issues for high stretch environments. Teledyne DALSA’s Marketing and Sales teams have approved it for up to 5 application size changes. Further development is required to increase lifetime “flex-ability”.
A major challenge in the design of image sensors is the need for long design cycles in order to adapt a sensor to a target application. The new flex sensor offers application flexibility previously unheard of. Not only can it expand to several times its size (one wafer scale device could easily cover the required area of a typical X-ray surgical panel) it can also be adjusted locally or non-symmetrically to become near-round, thereby adapting to distortions in the scene or optics.
Ultimately our roadmap will lead to a single device that can be customized on-site by stretching it over a provided mounting frame, then trimming it to size with a utility knife. Any alignment errors will be compensated for by leaving a few additional rows or columns on the device and windowing to the required size electronically. We recommend stretching the device to its maximum capability to avoid cutting partial pixels.
This adjustment method is supported by the negative EEC (Electro-Expansion-Coefficient) of the aluminume. When a positive voltage is applied to the bus lines they contract slightly into the base material at the cut edges and thus prevent loose ends and possible corrosion or shorts. This requirement prevents the sensor engineers from using negative biases on chip. Our CMOS team has found this to be just a small hurdle in their way to achieving excellent image quality and sensor performance.
From the Director of Manufacturing: “When I first heard about the concept, I thought ‘the CMOS designers are nuts – we use this substrate on the clean room floor to attract particles – what do they think this will look like on a sensor?’ – but the fully-cured flex sensor is quite resilient and, in fact, has solved a number of previous wash-down issues with normal sensors. We basically trim it, run it through the quick-wash cycle of our dedicated dishwasher and – voilá, ready to ship!”
The final device will offer a high degree of flexibility and a high number of flex-counts. Applications can fully surround themselves with vision, as the material can be enclosed by the image sensor and a 1:1 resolution ratio between the object and pixel can be achieved. With the use of transparent silicone, light can be applied through the sensors back side, making BST (back side thinning) a thing of the past. This material can also be dyed. Having already demonstrated uniform or patterned color mosaics, Teledyne DALSA is investigating a tie-dye pattern to allow spherical spectral separation for spectronomy applications. And just because it looks groovy!
The company’s Lead Technology Officer had this to say, “The color aspects of the new flex sensor may finally allow us to complete the quest for purple photons, something that has evaded the imaging community for decades.”