Abstract:
A hot spare boot circuit that automatically switches from a non-operational CPU to an operational CPU for powering up the computer system. In the multiprocessor computer system, a first CPU is designated to perform power on operations. If the first CPU fails, which is determined when a dead man counter in the hot spare boot circuit times out, the hot spare circuit ensures that the first CPU is in a disabled state. Next, the hot spare boot circuit identifies an operational second CPU, reinitializing certain ID information as necessary such that the second CPU can properly perform power on operations. The hot spare boot then awakens the second CPU, using a startup interprocessor interrupt in one embodiment, or simply negating the hard reset of the second CPU in a second embodiment. The second CPU then proceeds to perform the power on functions.
Abstract:
A power supply monitoring and control circuit using a microcontroller to remotely monitor and control the functions and conditions of a power supply. The power supply monitoring and control circuit is coupled to the primary and secondary sides of a power supply to monitor important voltage and current signals of the power supply, such as the output voltages and currents, and to control the various parameters of the power supply such as the output voltage and current limits. Analog to digital interface circuitry is provided to convert the power supply voltage and current signals to digital signals which are retrieved by a microcontroller which converts the digital signals to numbers representing the values of the power supply signals, and then stores the numbers. The microcontroller is also interfaced to reference and feedback signals of the power supply to control the power supply's operation. The microcontroller further keeps track of the total elapsed time of operation and the total number of times the power supply has been powered up. The power supply monitoring and control circuit operates as a slave to a host computer system so that a system operator can retrieve all of the monitored information and can control the operation of power supply. The host computer system communicates with the power supply monitoring and control circuit through a serial link, so that the host computer can be remotely located. The present invention also provides self-calibration to assure accurate data.
Abstract:
A memory controller which makes maximum use of any processor pipelining and runs a large number of cycles concurrently. The memory controller can utilize different speed memory devices at their desired optimal speeds. The functions are performed by a plurality of simple, interdependent state machines, each responsible for one small portion of the overall operation. As each state machine reaches has completed its function, it notifies a related state machine that it can now proceed and proceeds to wait for its next start or proceed indication. The next state machine operates in a similar fashion. The state machines responsible for the earlier portions of a cycle have started their tasks on the next cycle before the state machines responsible for the later portions of the cycle have completed their tasks. The memory controller is logically organized as three main blocks, a front end block, a memory block and a host block, each being responsible for interactions with its related bus and components and interacting with the various other blocks for handshaking. The memory controller operates in system management mode to override any write protect status of memory so that the SMRAM can be located in the main memory space and be write protected during normal operations but be full usable during system management mode.
Abstract:
A computer system determines the presence of a device on the system bus that responds to I/O or memory reads by performing an I/O or memory read with data bus pulled to its normally undriven state. If the value returned is other than the data bus' normally undriven value, it is determined that a device is responsive to that I/O or memory read. Otherwise, the system then pulls the data bus to other than its normally undriven state and performs another I/O or memory read. If the value returned is again the value of the data bus' normally undriven state, it is determined that a device is present because it is driving the data bus back to its normally undriven state. Otherwise, it is determined that a device is not driving the data bus in response to an I/O or memory read. Further, the computer system according to the invention makes the determination of whether a device is driving the bus using a comparator to compare the level present on the data bus in response to the I/O or memory reads.
Abstract:
A boot strap control circuit including a depletion mode NMOS transistor to effectively switch a boot strap bleed resistor out of the circuit after power up is achieved. The NMOS transistor is initially turned fully on to allow current to flow through the bleed resistor to a pulse width modulation circuit (PWM) upon power up and to allow early control by the PWM. When the PWM reaches operating power, it asserts its reference voltage output high, activating a transistor switch to turn the NMOS transistor fully off. A capacitor maintains power to the PWM until an auxiliary winding of the power transformer develops sufficient voltage to operate the PWM. In the event of failure of the auxiliary winding, the NMOS transistor operates discretely, turning on and off at a low duty cycle, rather than operating in the linear mode. In this manner, the NMOS transistor and bleed resistor assure low power during normal operation and safe operation at all times.
Abstract:
A circuit that automatically detects whether an input/output expansion board is connected to an EISA system or an ISA system. The circuit monitors the expansion bus for EISA slot-specific I/O cycles by sampling the bus signals AENx and BALE when either of the IORC, or IOWC, signals are asserted. When the circuit detects an EISA slot-specific I/O operation on the expansion bus, a signal is generated indicating that the expansion bus is the EISA bus. This determination allows an ISA expansion board to take advantage of certain EISA features when it is connected to an EISA system. Using a circuit to determine expansion bus type removes the need for a jumper to provide the same function, thus providing greater ease of use.
Abstract:
A tapped inductor slave regulating circuit provides a second slave output voltage derived from a tapped connection to the filter storage inductor of a first output voltage of a switching power supply converter. In the converter, an unregulated voltage is provided through a switching circuit to a storage inductor to develop a first output. The switching circuit is turned off and a synchronous rectifier is turned on to freewheel the current through the storage inductor and the load. The storage inductor is center-tapped and coupled to a switching circuit to provide a second slaved output. The location of the center tap is chosen to provide the proper voltage of the second output. In one embodiment, the switching circuit for the slaved output is turned on during the freewheel portion of each cycle to provide a proper voltage level for the second output. In another embodiment, a separate local feedback circuit is provided to further regulate the second output voltage level.
Abstract:
A synchronous memory controller capable of operating with three different frequency microprocessors and yet providing similar DRAM timings. Input frequencies of 32, 25 and 33 MHz correspond to 16, 25 and 33 MHz microprocessors. Various states are bypassed at certain frequencies to allow the various memory, latch and buffer control signals to be produced uniformly. The memory controller also handles operations from external buses, such as the EISA and ISA buses at the various input frequencies. These external bus cycles are controlled by separate state machines, which also have states bypassed for certain input frequencies.
Abstract:
A computer system has a processor coupled to a cache controller, uses page mode memory devices and performs page hit detection on the processor local bus. Column address and data values are latched by a memory controller on memory write operations to allow early completion of the cycle so that the next cycle can partially overlap. This allows the use of economical memories and yet have zero wait state page hit operation.
Abstract:
At least two double buffers or FIFOs, each FIFO including a first group of latches in series with a second group of latches, coupled between a host data bus and a corresponding bank or way of interleaved memory. The inputs of the first group of latches are coupled to the host data bus, and the inputs of the second group of latches are coupled to the outputs of the first group of latches of each double buffer or FIFO. The outputs of the second group of latches are coupled to the memory data bus of the corresponding way of interleaved memory. During a burst write sequence, an address is placed on the host address bus and a series of data doublewords are sequentially placed onto the host data bus, while the DRAMs of main memory are entering into page mode. The first group of latches of each double buffer or FIFO latches in every other data doubleword. The second level of latches stores the data from the corresponding first level of latches to provide the data to the DRAMs according to the timing requirements of the DRAMs. In this manner, the CPU or cache controller providing data to the host bus may operate at full speed without inserting wait states while the DRAMs enter into page mode.