Camera Options for the Biomedical Instrument Builder
There are many considerations when selecting an imaging system for quantitative, low-light imaging. We work with our OEM customers to educate their teams on the choices – to ensure the best selections are made to meet performance and cost targets. Our support also goes beyond camera options to include advice on mounting and placement, thermal management, optical design and image related software.
When the instrument needs to fit a tight budget, we can help you get the most from routine cameras, ensuring you are not paying for capabilities that are not needed. And when every photon of signal is required, we have the highest performing back-illuminated EMCCD cameras available. Photometrics and QImaging understand the importance of the selection process, and will help you avoid cookie cutter solutions.
Sensor cooling is used to eliminate noise resulting from thermal background signal, known as dark current, an intrinsic property of semiconductor detectors. The amount of thermal noise that can be tolerated depends on the application requirements and image sensor. In camera selection, the actual dark current per exposure time is far more important than the specified temperature, as the true sensor temperature can vary widely. In nearly all cases, camera sensitivity and measurement precision can be improved by reducing thermal noise through use of well regulated, Thermo-Electric Cooling (TEC), typically resulting in a thermal noise of few electrons per pixel or less for a given exposure time.
TEC coolers offer a simple, compact solution, but are electrically inefficient, generating far more heat then they remove. For this reason, cameras offer several choices for removing heat from the TEC. Our OEM products offer a selection of passive air (limited heat removal but highly reliable, convenient and without vibration), forced air using low vibration fans (increased heat removal, convenient, but reduced long-term reliability) and liquid cooling (powerful, low vibration, but adds cost and maintenance).
In all cases, the amount of heat to remove is reduced by thermally isolating the sensor. This can be as simple as mechanical isolation with the sensor in an inert gas with reduced thermal conductivity, to as complex as a permanently sealed, high vacuum environment that eliminates convective heat loads all together.
Photometrics and QImaging offer a wide range of sensor technologies, sensor area and pixel sizes. Often, changing between sensor variants within a given family is easily done, so the list of available sensors is far larger than shown.
Current sensor technologies are Electron Multiplication CCDs (EMCCD), Scientific CMOS, Standard CMOS, as well as Interline, Frame Transfer and traditional Full Frame CCDs. Several of these sensors are available in front and back illuminated versions. Sensor manufacturers frequently used by Photometrics and QImaging are OnSemi (formerly Kodak), Sony, E2V, BAE and other custom sensor fabs.
Recent advances in FPGA and GP-GPU signal processing have enabled a significant increases in the real-time processing power of Photometrics and QImaging cameras. Recent cameras, including the new QI OEM series, perform defective pixel correction, removal of dark current background and implement HDR algorithms, all on camera using an FPGA (field programmable gate arrays with specialized signal processing capabilities). This processing power can also be extended to embedding OEM’s own algorithms inside the camera, in an IP protected environment. In addition, instrumentation builders can leverage in-line parallel processing on the host PC using CUDA and OpenCL based GP-GPU processing. To simplify the integration of streaming pixel data and GP-GPU processing, the PVCAM SPLICE library is available to instrument builders.
Host Computer Interfaces
Photometrics’ experience in building digital cameras predates PC technology, while QImaging was the first to introduce the convenience of FireWire interfaces. This deep experience has provided lasting insights on how to maximize performance across variable PC architectures without increasing cost.
PVCAM cameras support USB2, USB3, IEEE1394, and PCIe host computer interfaces, under Windows 7 (32 bit and 64 bit), Windows 8 (64 bit), Linux (32 bit and 64 bit) on Intel processors, and Linux (32 bit) on ARM processors. The host computer is chosen based on the applications and markets served by the camera, and in some cases the camera offers the choice of interface depending on performance.
PVCAM (Photometrics Virtual Camera Access Method) is a standard API across Photometrics and QImaging cameras. PVCAM uses a smart camera philosophy, where the API queries the camera for its capabilities. This not only simplifies the library, it allows the software architect to write general interfaces, and still easily switch between camera models without rewriting applications, even across different technologies like CCD and sCMOS. The benefit is that instrument builders can start development well before a particular camera is available, or quickly switch camera models when markets shift, without significant rework. This approach has been so successful that it is now widely mimicked across the industry, but it seldom achieved without a true intelligent camera approach.
3D Design and Modeling
The speed of current design requires full attention to 3D design practices. The process starts with sharing 3D models (STEP files are available on the product pages in this site) to optimize camera location and mounting points. Advanced CFD modeling is used not only for sensor cooling design, but also to assist OEMs to ensure camera operation is maintained inside their instrument enclosure.