Browse Prior Art Database

Synchronizable Window Watchdog

IP.com Disclosure Number: IPCOM000099119D
Published in the IP.com Journal: Volume 5 Issue 4 (2005-04-16)
Included in the Prior Art Database: 2005-Apr-16
Document File: 2 page(s) / 218K

Publishing Venue

Siemens

Related People

Juergen Carstens: CONTACT

Abstract

Watchdog functions are used to ensure that system microprocessors or controllers do not operate erroneously if the power supply voltage is outside the specified operating range, which is especially critical during system power up and power down sequencing. This is usually achieved by introducing certain preset time windows during which the watchdog (WD) expects a signal from the device to be monitored (Open Window) and others where there should be no signal (Closed Window). If the watchdog does not receive the test signals in the correct order, a warning is issued and/or a reset is induced. Problems arise when the available time base of the device to be checked is not stable during the start-up or the time until it becomes stable differs considerably depending on the situation (e.g. temperature) at the start-up. In particular, it has so far not been possible to include a function check of the watchdog itself without introducing a more stable and therefore more cost intensive time base. In the following a method is proposed that allows for a manual start of the WD (e.g. by the microcontroller) instead of an automatic start-up with a fixed timing as it was typical until now. In addition, a latch of the watchdog state allows for a functional test of the WD. This is done by applying a timing failure to check whether the watchdog notices the error. A typical embodiment of the proposed watchdog is depicted in Figure 1; the state diagram is shown in Figure 2. After a reset, the WD waits for a trigger signal by the microcontroller (Wait Window; WW) via the watchdog input (WDI). The internal Wait_Latch (WL) is set to 'Low'. As soon as this trigger signal is detected, the system proceeds to the 'Closed Window' state (CW) and, after a preset 'Closed Window Time' (tCW), the watchdog goes on to the 'Open Window' state (OW). A trigger signal within CW is interpreted as a pre-trigger failure and results in a 'Low' signal at the WD output pin. If a valid trigger signal is detected during the OW, the CW is initialized immediately. Otherwise, it returns to the CW after a preset 'Maximum Open Window Time' (tOW).

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

S

Synchronizable Window Watchdog

Idea: Alberto Flore, IT-Padova; Marco Piselli, IT-Padova; Christian Schweikert, DE-Munich

Watchdog functions are used to ensure that system microprocessors or controllers do not operate erroneously if the power supply voltage is outside the specified operating range, which is especially critical during system power up and power down sequencing. This is usually achieved by introducing certain preset time windows during which the watchdog (WD) expects a signal from the device to be monitored (Open Window) and others where there should be no signal (Closed Window). If the watchdog does not receive the test signals in the correct order, a warning is issued and/or a reset is induced.

Problems arise when the available time base of the device to be checked is not stable during the start- up or the time until it becomes stable differs considerably depending on the situation (e.g. temperature) at the start-up. In particular, it has so far not been possible to include a function check of the watchdog itself without introducing a more stable and therefore more cost intensive time base. In the following a method is proposed that allows for a manual start of the WD (e.g. by the microcontroller) instead of an automatic start-up with a fixed timing as it was typical until now. In addition, a latch of the watchdog state allows for a functional test of the WD. This is done by applying a timing failure to check whether the watchdog notices the error.

A typical embodiment of the proposed watchdog is depicted in Figure 1; the state diagram is shown in Figure 2. After a reset, the WD waits for a trigger signal by the microcontroller (Wait Window; WW) via the watchdog input (WDI). The internal Wait_Latch (WL) is set to 'Low'. As soon as this trigger signal is detected, the system proceeds to the 'Closed Window' state (CW) and, after a...