Cheetah PCB
0-260kfps AI-in-Sensor Imager
Cheetah's Revolutionary Capabilities
Cheetah PCB is an AI-in-Sensor imager capable of 0 to 260,000 frames per second (fps) operation. Cheetah is available in two variants: Visible range or extended IR operation (up to 950nm). The integration time and frame rate are programmable: At 80×120 pixels, the frame rate is up to 40k frames/sec; at 80×60 or 40×120, the frame rate is up to 77kfps; at 40×60 the frame rate is up to 134kfps, and at 40×30 the frame rate is up to 260,000 frames per second. Each pixel is converted to an 8-bit digital representation and output via proprietary parallel interfaces. Cheetah also includes an LED driver capable of up to 40mA continuous current (programmable) and PWM control. Multiple registers may be set using an SPI interface, including exposure time, interval between exposures, multi-picture exposure count or direct shoot, LED current, PWM duty cycle, etc. For higher current levels an external transistor may be used to buffer the current capability. All blocks required to create a high-speed camera are included including timing, power management, imaging, LED driver, and four configurable GPIO’s capable of up to 100MHz operation
Key Features
20×44 Ultra High Speed Imager
– up to 10kFPS
Charge-Sensor Pixel Imager
– Direct pixel charge processing
– Charge to digital conversion
On board power management
4 High speed 8 bit parallel digital I/Os for pixel data output (32 I/Os total)
SPI Interface
On board VCO and PLL
72 pin OLGA 6.4×6.4mm
LED driver supporting internal and external PWM control (up to 40mA)
PWM driver for generating external negative bias voltage
Four configurable GPIO’s
Applications
Gamer Mouse Frontend · High Speed Tracking · High Speed Inspection ·
High Speed Assembly · Bar Code Readers · Semiconductor Inspection · PCB Inspection
Cheetah – Golf Application
Ultra High-Speed Cheetah watching a golf ball fly by and a facet on the front of a golf club fly by. Using Cheetah, customers can use the speed and angle information to create images to predict where the ball is going to go through the end user’s cell phone. All of this would be possible with 5kfps and at a fraction of our competitions cost.
