This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

SM320F28335-EP: Watchdog Timer

Part Number: SM320F28335-EP

Hello,

I'm using the TI SM320F28335GJZMEP microcontroller along with the TI voltage supervisory watchdog timer TPS3823-33DBVR.  The circuitry I'm using is shown below

From the TPS3823-33DBVR datasheet, if WDI remains high or low longer than 1.6 seconds, then reset is triggered.  The problem I'm having is that when I apply power to the MCU (before I can program the MCU to provide the WDI pulse heartbeat), the watchdog timer TPS3823-33DBVR pulls *RESET low, causing the MCU to go into reset mode, which prevents me from being able to program over JTAG to provide the WDI pulse heartbeat.

How can I get around this problem?  How are watchdog timers typically used to not have this issue?

Thanks,

Kevin

  • Kevin,

                    Let me start by first saying that TPS3823-33DBVR is not the right choice, since it has a push-pull output. Remember that the -XRS pin is bi-directional. i.e. it can also be driven by the DSC upon an overflow of the internal watchdog counter.  It is for this reason our datasheet recommends a OD device: "The output buffer of this pin is an open-drain with an internal pullup. It is recommended that this pin be driven by an open-drain device". (on a slightly different note, I recommend you refer to www.ti.com/lit /SPRS439 for all non EP-device-specific information, since SPRS581 has not been updated in a while).

    Right now, the only choice you have is to remove R48, program the device and put R48 back in the circuit. I don't know if you have a R-C circuit connected to the -XRS pin. If not, I am not sure if the device would get a proper reset.

  • Hareesh,

    Thanks for your reply, I understand now that the MCU datasheet recommends connecting "-XRS" to an open drain and that the -XRS can also be driven by the MCU upon an overflow of the internal watchdog counter.

    However, would connecting to an open drain supervisory IC resolve the problem I am having? 

    An open drain supervisory IC would still pull  -XRS low in the absence of a WDI pulse (there would be no WDI pulse until I could program it), which would cause the MCU to be constantly going into a reset mode preventing me from programming it.

    Also the link you put in your reply is broken.  Is there a new link for that app note?

    Thanks,

    Kevin

  • However, would connecting to an open drain supervisory IC resolve the problem I am having? 

    Unfortunately, no.

    An open drain supervisory IC would still pull  -XRS low in the absence of a WDI pulse

    True.

    (there would be no WDI pulse until I could program it), which would cause the MCU to be constantly going into a reset mode preventing me from programming it.

    A catch-22 situation. The only solution I can think of for your specific situation is to have a series jumper (or a DIP switch) that would disconnect the TPS device when you want to program the flash. Again, this is assuming there is a R-C circuit on the -XRS pin that would reset the device properly upon a power-up.

    Also the link you put in your reply is broken.

    www.ti.com/lit/SPRS439

  • A catch-22 situation. The only solution I can think of for your specific situation is to have a series jumper (or a DIP switch) that would disconnect the TPS device when you want to program the flash. Again, this is assuming there is a R-C circuit on the -XRS pin that would reset the device properly upon a power-up.

    Watchdog timers seem like they are pretty widely used from what I've read online.  Are watchdog timers typically implemented with a series jumper/DIP switch to allow for programming?  Is my specific situation different from how watchdog timers are typically used?

  • Watchdog timers seem like they are pretty widely used from what I've read online. 

    Agree, but designers typically rely on the on-chip WD, not an external WD.

    Are watchdog timers typically implemented with a series jumper/DIP switch to allow for programming? 

    For your specific design, that is the only solution I am able to think of.

    Is my specific situation different from how watchdog timers are typically used?

    As mentioned earlier, I don't recall seeing an external WD being used and in this way. The current implementation may be OK if the firmware is already programmed when the DSP is soldered on the board and the application will never require a firmware update. Unfortunately, it neither allows initial programming of the Flash nor future updates.

  • A catch-22 situation. The only solution I can think of for your specific situation is to have a series jumper (or a DIP switch) that would disconnect the TPS device when you want to program the flash. Again, this is assuming there is a R-C circuit on the -XRS pin that would reset the device properly upon a power-up.

    Would it be possible to do something similar to as described in Section 3.5.2 of

    https://www.ti.com/lit/ug/slau533d/slau533d.pdf?ts=1670978726333&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FMSP-EXP430F5529LP

    with a pre-initialization function that disables the reset or does that only work for internal watchdog timer?

    Thanks,

    Kevin

  • We do have something similar for C2000 devices as well, but only for the internal WD. There is no way to disable the reset function of the -XRS pin.