It is no secret that cell phones are cheap and small. When you add consumer demand for more functionality that includes higher-resolution imaging for stills and videos, you have a rather interesting case study in the classic engineering discipline of compromise.
CMOS image sensors, which have always had to battle CCDs at the high end in terms of resolution and image quality, have a distinct advantage in these applications. The imagers can be run off on existing CMOS fab lines for low-cost production, and they can integrate a wide variety of image processing, compression and other logic circuits to streamline designs for small size.
To get a better idea of the options, let's take a look at the basic features required for camera phones and their related applications. These include quality images, demosaicing (translating the Bayer color filter bit stream into a real image), noise compensation, auto white balance, auto exposure, small file size and low power.
The standard for reducing file size and memory requirements is JPEG compression. This algorithm appears on some image sensor ICs but is most often handled by a companion chip known as the image signal processor, or ISP. Before the advent of the CMOS image sensor (CIS), all of the image processing took place on a devoted processing IC.
While that approach is still true today for both charge-coupled device (CCD) and CIS sensors, it has begun to change. As a result, solid-state imaging has been dropped into the system-on-chip vs. system-in-package battleground. The remaining features are the minimum needed to create images of reasonable quality.
Developers that want to put digital processing onto the image processor focus on the most obvious issue: overall camera footprint. Avago is a leader in the integration of digital processing onto the sensor die. In fact, Avago is the only vendor with a CIS device that in- cludes on-chip JPEG encoding and autofocus control. Naturally, it advocates the use of a single system-on-chip (SoC) device in the camera module.
In an EE Times story by Patrick Mannion last October, Avago marketing manager Feisal Mosleh said, "The way the camera modules are mounted in handsets, they impact the thickness of the handset, which has become a major selling point. There just isn't the height available for a stacked package."
Certainly, handset thinness is a prime marketing feature, and OEM requirements for packages, components and modules restrict vendor physical designs. Despite the argument that component thickness is critical and leaves no room to spare, it is important to recognize that the thickness under discussion is really that of a single die. Although there is some additional overhead related to die attachment, spacers and wire bonding, die thickness for an ISP can be 0.2 mm or less. It will be up to each module vendor to decide if this is a significant percentage of the module height. For the Standard Mobile Imaging Architecture standard, the maximum height is 7.6 mm.
To put the module height restrictions into a consumer context, some "ultrathin" handsets that have been introduced include Samsung's SGH-P300, the "credit card phone," which is only 9 mm thick. The Motorola Razr v3i (clamshell) is 14 mm, Samsung's t509 is 9.8 mm thick and its Xcute S50 is 9 mm thick.
