Distributed diode fault check

• US 6,967,487 B2
• Filed: 06/04/2003
• Issued: 11/22/2005
• Est. Priority Date: 06/06/2002
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Garnett, Paul J.
• Primary Examiner(s): Nguyen, Vincent Q.
• Application Number:
  US10/455,753
• Publication Number:
  US 20040042145A1
• Time in Patent Office: 902 Days
• Field of Search: 324/158.1, 324/767, 324/537, 340/635, 340/645, 340/650, 340/652, 340/657, 340/660, 340/664, 340/514
• US Class Current: 324/537
• CPC Class Codes: G01R 31/30 Marginal testing, e.g. by varying supply voltage testing computers during standby operation or idle time G06F11/22
• CPC Class Codes: G01R 31/40 Testing power supplies testing photovoltaic devices H02S50/10
• CPC Class Codes: G06F 1/28 Supervision thereof, e.g. detecting power-supply failure by out of limits supervision
• CPC Class Codes: G06F 11/2015 Redundant power supplies power supply failure G06F1/30
• CPC Class Codes: H02J 1/108 using diodes blocking reverse current flow H02J1/12 takes precedence

Abstract

A method and apparatus for testing for latent faults in the isolation devices of a system including redundant power supplies which supply power to one or more system units. A system controller is operable to perform a test cycle to perform the fault checks, which may including checks for short circuits and/or open circuits. The checks may be performed within a test cycle in which each of the isolation devices in the system is tested. The test cycle may be performed at regular intervals, the interval between each cycle being determinable by user input. In the event that a fault in one of the isolation devices is detected, the system controller may be operable to report the fault to an alarm system.

Claims

1. A fault test method for use in a system comprising one or more system units, wherein each system unit comprises at least two isolation devices and is powered by a plurality of redundant power supplies, each isolation device requires a forward biasing voltage at least equal to a respective forward bias voltage value to be operable in a normal conductive state and is electrically connected with a respective one of said power supplies such that each power supply has associated with it a set of one or more isolation devices, and an output of each isolation device is connected to a common load of the corresponding system unit, the method being performed by a testing system and comprising:

selecting as a subject for a test procedure, the set of isolation devices associated with a corresponding one of said power supplies;

controlling the power supplies during the test procedure to marginally vary an output voltage of at least one power supply such that a differential voltage is applied between an output of the power supply associated with the selected set of isolation devices and outputs of the remaining power supplies for selectively changing the conductive state of at least one isolation device by reducing the forward voltage thereof to less than said respective forward bias voltage;

measuring the current through each isolation device in said selected set of isolation devices; and

determining whether each measured current is consistent with the existence of a fault condition in the respective isolation device.

2. The fault test method of claim 1, wherein said test procedure includes a short circuit detection test in which each of the power supplies other than the power supply associated with the selected set of isolation devices has a first output voltage greater than a second output voltage of the power supply associated with the selected set of isolation devices such that the differential voltage is applied for changing the conductive states of each isolation device of the selected set of isolation devices to a non-conductive state in the absence of a fault condition and wherein, if the measured current through any isolation devices of the selected set of isolation devices is non-zero and in a direction towards the corresponding power supply, the measured current is determined to be consistent with a short circuit fault condition in that isolation device.

3. The fault test method of claim 2, wherein said test procedure includes an open circuit detection test in which the power supply associated with the selected set of isolation devices has a first output voltage greater than a second output voltage of the remaining power supplies such that the differential voltage is applied for changing the conductive states of each of the isolation devices in sets other than the selected set of isolation devices in the absence of a fault condition and wherein, if the current measured through any of the selected isolation devices is substantially zero, the measured current is determined to be consistent with an open circuit fault condition in that isolation device.

4. The fault test method of claim 3, wherein the method comprises performing a test cycle comprising a sequence of test procedures such that in a completed test cycle respective short circuit tests and open circuit tests are performed for each of the isolation devices.

5. The fault test method as claimed in claim 4, wherein a system controller initiates a test cycle at predetermined intervals.

6. The fault test method as claimed in claim 5, including the system controller receiving and storing input data representative of the predetermined intervals and initiating the test cycle at intervals determined by the data.

7. The fault test method as claimed in claim 5, comprising the system controller generating an alarm signal in response to detection of a fault condition.

8. The fault test method of claim 5, wherein one of said system units comprises said system controller.

9. The fault test method of claim 8, wherein the common load of the system unit comprising said system controller comprises the load presented by said system controller.

10. The fault test method of claim 5, comprising the system controller outputting command signals for controlling the power supplies, each isolation device having a respective sensing circuit transmitting signals representative of the measurement of current to the system controller;

and the system controller determining whether the measured currents are consistent with the existence of a fault condition.

11. The fault test method of claim 1, wherein the differential voltage is applied such that, during the test procedure and in the absence of fault conditions, the common load of each system unit is supplied by at least one of the power supplies operating at a first output voltage and being associated with a respective set of isolation devices each in a normal conductive state, and such that each remaining power supply operates at a second output voltage by the differential amount and in respect of which power supply the resulting forward voltage across each of its respective set of isolation devices is less than the forward bias voltage level of that isolation device so that conduction is inhibited.

12. The fault test method of claim 11, wherein said differential voltage is applied such that during the test procedure and in the presence of a fault condition requiring each common load to be supplied by one or more of the remaining power supplies operating at the second voltage, each resulting common load voltage is greater than a minimum operational limit required for operation of that common load.

13. The fault test method of claim 1, wherein each power supply comprises a respective margin arm input and wherein the controlling step comprises inputting a respective logic level signal to the margin arm input of the each power supply.

14. The fault test method of claim 1, wherein each isolation device comprises a respective diode.

15. The fault test method of claim 1, wherein said system comprises a computer system.

16. The fault test method of claim 1, wherein at least one system unit comprises a field replaceable unit.

17. The fault test method of claim 1, wherein each set of isolation devices is powered by the corresponding power supply via a common power bus.

18. A carrier medium carrying processor implementable instructions for instructing a processor of a testing system to perform a fault test method for use in a system comprising one or more system units, wherein each system unit comprises at least two isolation devices and is powered by a plurality of redundant power supplies, each isolation device requires a forward biasing voltage at least equal to a respective forward bias voltage value to be operable in a normal conductive state and is electrically connected with a respective one of said power supplies such that each power supply has associated with it a set of one or more isolation devices, and an output of each isolation device is connected to a common load of the corresponding system unit, the method comprising:

selecting as a subject for a test procedure, the set of isolation devices associated with a corresponding one of said power supplies;

controlling the power supplies during the test procedure to marginally vary an output voltage of at least one power supply such that a differential voltage is applied between an output of the power supply associated with the selected set of isolation devices and outputs of the remaining power supplies for selectively changing the conductive state of at least one isolation device by reducing the forward voltage thereof to less than said respective forward bias voltage;

measuring the current through each isolation device in said selected set of isolation devices; and

determining whether each measured current is consistent with the existence of a fault condition in the respective isolation device.

19. A system comprising:

one or more system units, each system unit comprising at least two isolation devices, and measuring circuitry for measuring the current through isolation devices, each isolation device requiring a forward biasing voltage at least equal to a respective forward bias voltage value to be operable in a normal conductive state and having an output connected to a common load of the respective system unit;

a plurality of redundant power supplies, each supplying power for at least one system unit via an isolation device of the respective system unit such that each power supply has associated with it a set of one or more isolation devices, each power supply comprising control circuitry operable to control an output voltage of that power supply; and

a system controller, operable;

to select as a subject for a test procedure, the set of isolation devices being associated with a corresponding one of said power supplies;

to cause at least one power supply to marginally vary an output voltage thereof such that a differential voltage is applied between an output of the power supply associated with a selected set of isolation devices and outputs of the remaining power supplies for selectively changing the conductive state of at least one isolation device by reducing the forward voltage thereof to less than said respective forward bias voltage, and

to determine whether the current measured by said measuring circuitry through each isolation device in said selected set of isolation devices is consistent with the existence of a fault condition in the respective isolation device.

20. The system of claim 19, wherein one of said system units comprises said system controller.

21. The system of claim 20, wherein the common load of the system unit comprising said system controller comprises the operational load of said system controller.

22. The system of claim 19, wherein the system controller is operable to initiate a test cycle at predetermined intervals.

23. The system of claim 22, wherein the system controller is operable to receive and store input data representative of the predetermined intervals and to initiate the test cycle at intervals determined by the data.

24. The system of claim 22, wherein the system controller is operable to generate an alarm signal in response to detection of a fault condition.

25. The system of claim 22, wherein the system controller is operable to output command signals for controlling the power supplies, each isolation device having a respective sensing circuit operable to transmit signals representative of the measurement of current to the system controller;

whereby the system controller is further operable to determine whether the measured currents are consistent with the existence of a fault condition.

26. The system of claim 19, wherein each power supply comprises a respective margin arm input and wherein the output voltage of that power supply may be varied by inputting a respective logic level signal to that margin arm input.

27. The system of claim 19, wherein said system comprises a computer system.

28. The system of claim 19, wherein at least one system unit comprises a field replaceable unit.

29. The system of claim 19, wherein each isolation device comprises a respective diode.

30. The system of claim 19, wherein each set of isolation devices is powered by the corresponding power supply via a common power bus.

31. The system of claim 19, comprising a rack mountable shelf, the rack mountable shelf supporting the plurality of power supply units and a plurality of said system units in the form of server blades.