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.

quetion about the diffrence between the diagnotic state and active state of TPS65381

Other Parts Discussed in Thread: TPS65381-Q1

hi,all

     I use TPS65381 as watchdog for TMS570, I want to stop watchdog function during work for some reason ,and I find I havo to put the TPS65381 in diagnotic state ,then I can disable the watchdog function , so I need knowing the diffrence  between the diagnotic state and active state of TPS65381.I found under  diagnotic state ,if stop feeding the watchdog ,it can generate reset signal ,why?

  • Hello:

    I don't really understand what is the use case to stop the watchdog once it was already started. I would recommend you to take a look at the datasheet (www.ti.com/.../tps65381-q1.pdf). I'll relocate your post to the corresponding group for further assistance.

    Regards,

    Enrique
  •       Thank you for your reply ,I have already read the pdf ,I feel the pdf  did'nt describe in detail,so I have to ask you for help.The reason why I want to stop watchdog is clear ,when I find someting wrong in the system ,I want to stop the output ,input and interrupt of the TMS570,but I don't want to be restarted ,you know ,if the interrupt is stopped ,the time interval can't be generated , so watchdog will restart the CPU.I use RTOS on the TMS570 ,so the interrupt is controlled by the RTOS,if I stop interrupt ,all interrupt is stopped.

          Hope to get answer soon .Thank you.

  • anther little problem , I click on your link ,but the web show"Sorry! We couldn't find your page."
  • Hi Jiaming,

    The WD timer RESET capability is enabled and disabled independently of the state of the TPS65381-Q1 device.   Enabling RESET from a WD failure (WD_FAIL_CNT increments to 7+1) is controlled by the WD_RST_EN bit in the SAFETY_FUNC_CFG register (bit D3).  If this bit is SET to 1 even in DIAGNOSTIC state and the watchdog counter sees too many "bad events" the device will cause NRES to go low and it will transition the TPS65381-Q1 through its RESET state as well back to DIAGNOSTIC state. 

    There are many differences between DIAGNOSTIC STATE and ACTIVE state as outlined in the datasheet which can be found in this product folder:  http://www.ti.com/product/tps65381-q1.

    The key differences and point to this specific scenario and question are:

    1) Watchdog configuration cannot be changed in ACTIVE state. In other words the window durations, selection of trigger mode or Q&A mode and enabling or disabling of RESET on a watchdog failure is only allowed to be changed in DIAGNOSTIC state.  Section 5.4.1.12 outlines how the watchdog fail counter will work with good and bad events, which count level impacts ENDRV output (assuming it was enabled via ENABLE_DRV bit) and NRES output (assuming RESET on WD fail was enabled via WD_RST_EN).

    2) Many other configuration registers are locked during ACTIVE state.  Table 5.5.6 outlines the SPI commands and also highlights in a single table those where a SPI write or update can be given only in DIAGNOSTIC state.  ACTIVE state is the state for normal option where the TPS65381 is monitoring the MCU (to the level it was configured and enabled), thus the MCU cannot change this monitoring on the fly or it defeats the independent monitoring so the configuration registers are locked.

    3) After the device starts up, goes through RESET state to DIAGNOSTIC state, the WD_RST_EN will be cleared (0) so it will not cause a RESET from WD failure until the software as set the WD_RST_EN bit.  The job of the WD is to cover the MCU so if RESET on WD failure has been set, that is what the device will do.  If you have other diagnostic functions that may stop feeding the watchdog during DIAGNOSTIC state they need to be run prior to configuring and synchronizing to the watchdog and enabling WD_RST_EN.

    Best Regards,

    Scott Monroe

  • thank you for your reply, I think if I lock the register ,the diagnotic state is nearly the same state of activate,isn't it?
  • From a certain viewpoint that is true, however since the MCU can still unlock the registers it could conceivably re-configure the TPS65381 when it shouldn't.  Going to ACTIVE state means a power cycle or RESET event has to happen to go back through the state machine to re-configure the device again.

    Scott

  • OK ,I understand.Thank you for your answer.