Abstract:
Analog-to-digital conversion through successive approximation is implemented by means of a charge coupled device. During the conversion process two charges are compared, each comparison yielding one bit of a multi-bit number. By increasing the lesser of the compared charges after each comparison, the need to subtract charge as part of the successive approximation process is eliminated.
Abstract:
Analog-to-digital conversion through successive approximation is implemented by means of a charge coupled device. During the conversion process two charges are compared, each comparison yielding one bit of a multi-bit number. By increasing the lesser of the compared charges after each comparison, the need to subtract charge as part of the successive approximation process is eliminated.
Abstract:
A space-saving back-up power supply apparatus has length and width dimensions substantially the same as those of a conventional integrated circuit connector. The apparatus includes sockets arranged on its top for making connection to a socket-pluggable integrated circuit such as a standard CMOS RAM, and the apparatus has pins extending from its bottom for making connection to a printed circuit board or connector of a host electronic system. Control circuitry and one or more batteries are located within the apparatus. The back-up power supply is operative to provide power to the socket-pluggable integrated circuit even if the normal power supply of the host electronic system is short-circuited. The control circuitry of the back-up power supply controls the chip enable signal and performs a battery test upon power-up.
Abstract:
A redundancy circuit that consumes no power before or after activation switches a pair of output nodes from a first set of complementary logic levels to an inverted set when it is activated by blowing a pair of fuses.
Abstract:
A dual storage cell memory includes an array of dual storage cells, each of the dual storage cells containing a first memory cell and a second memory cell. The first and second memory cells are well known six-transistor static memory cells with the addition of transfer circuitry for transferring data directly from the internal data nodes of each of the memory cells to its corresponding complementary memory cell without requiring the use of the enable transistors or the bit lines associated with each of the dual storage cells.
Abstract:
An embedded framing bit pattern in a serial bit stream is located by combining the last bit to arrive of the serial bit stream with a predetermined number of prior bits of the serial bit stream which are spaced apart by the pitch of the bits of the framing bit pattern, and this combination of the bits is tested to determine if the combination matches part of the framing bit pattern. If a match does not occur, then the bits which were combined together are changed to a bit pattern that will not result in a match when these bits (except for the eldest bit which is disregarded) is combined again with a new bit of the serial bit stream, no matter what the logic state of the new bit. In this manner all of the bits, as they arrive and are combined and tested, will eventually be changed except the bits which are part of the framing bit pattern.
Abstract:
A low-power crystal-controlled CMOS oscillator wherein a long and wide additional transistor is provided in the first stage of the output amplifier. This prevents the output amplifier from diverting too much current from the primary amplifier stage during startup.
Abstract:
A temperature and processing compensated time delay circuit of the type which can be fabricated in a monolithic integrated circuit utilizes a field effect transistor (FET) (12) connected to the terminals of a charged capacitor (14). A bias voltage connected to the gate of the FET (12) varies with temperature in a manner to compensate for the changes in current which flows from the capacitor (14) through the FET (12) due to changes in temperature. The bias voltage also varies from one integrated circuit to another in a manner to compensate for variations in FET threshold voltage caused by variations in the processing of the integrated circuits.
Abstract:
An embedded framing bit pattern in a serial bit stream is located by combining the last bit to arrive of the serial bit stream with a predetermined number of prior bits of the serial bit stream which are spaced apart by the pitch of the bits of the framing bit pattern, and this combination of bits is tested to determine if the combination matches part of the framing bit pattern. If a match does not occur, then the bits which were combined together are changed to a bit pattern that will not result in a match when these bits (except for the eldest bit which is disregarded) is combined again with a new bit of the serial bit stream, no matter what the logic state of the new bit. In this manner all of the bits, as they arrive and are combined and tested, will eventually be changed except the bits which are part of the framing bit pattern.
Abstract:
A data compression/decompression processor implements a modified Ziv-Lempel ("LZ") coding technique. The processor includes three modules, an interface, a coder-decoder ("CODEC"), and a MODEL. The CODEC and the MODEL modules together form compression engine, in which the CODEC provides variable length coding and data packing, and the MODEL implements the LZ processing. The MODEL uses content addressable memory ("CAM") in encoding mode for text storage and character matching, and uses CAM in decoding mode as an on-chip RAM to obtain high speed access.