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.

TIDA-00489: Looking for an external watchdog

Part Number: TIDA-00489
Other Parts Discussed in Thread: CC1310, TPS3431, TPS3813, TPL5010

Hey guys,

we use the TIDA 00489. We noticed that our system sometimes crashes after a months of operation in the field and we were not able to trace the bug, though we assume it might be externally induced (vibration, mechanical stress, EMV). We want to use an automatic reset system without the necessity of user interference. When we open the sensor housing, we can short circuit the sensor to restart the system and the sensor works fine again, hence, we are thinking about implementing an external watchdog to detect such lockups and restart the CC1310. The watchdog should fullfil the following requirements:

  • Cost around 50Cents
  • Low power applicability (below 5µA). It should run with the 32kHz clock already running on the TIDA pcb.
  • Watchdog window should be huge since it does not have to be fast, checking the system every 3hours would be sufficient
  • Should be running on 3.3V (better between 2.0 - 3.7V)
  • Small footprint with few peripherals

We actually really need an easy system. A watchdog waiting for a high edge every 2-3hrs and if not pulling the RST button of the CC1310 to GND would be sufficient. Something like a 5 pin chip with GND, Vcc, Clock Input, 1x input from CC1310 and 1x output connected to RST. The latter one driving RST to GND if necessary.

Best wishes

Slev1n

  • Slev1n, hello and good day.  Your query has been forwarded to the owner of this reference design.  Unfortunately, he is current on vacation and will not return to the office until Thursday.  Please be advised that upon his return to office, he will both review and respond to your question on Thursday, August 27th. 

    Best Regards,

    John Fullilove

    Reference Design Operations

    Texas Instruments

  • Thank you very much for the fast answer and hint. Thursday will be fine.

    Best wishes

    Slev1n

  • Slev1n,

    There are a couple different solutions available in IC form. 

    The first are our watchdog / supervisor devices.  A couple of these to take a look at would be the TPS3431 and the TPS3813.  Both of these are in your price range with the TPS3431 in a small 8 pin VSON package and the TPS3813 in a 6 pin DBV package.  The down side to these parts is the quiescent current which is ~10uA.

    The second solution I found was to use our Nanotimers for this function.  The TPL5010 appears to be a good device for this application.  The quiescent current is in the 10's of nA range, is half the cost of the supervisors above and comes in a SOT23 package.  It appears that the timer can be set as long as 2 hours on this particular device.  Section 8.3 on page 9 of the datasheet described how to use this device as a watchdog that closely matches your description.

    There are likely other solutions available.  These are the two which look to be the most promising to meet your specs.  Let us know if either of these work for you.

    Regards,

    David

  • Hey David,

    thank you for your search. I will check them and report back which one we will try.

    best wishes

    Slev1n

  • Hey David,

    I think the TPL5010 is what we are looking for. I've been through the application note and I have a few questions regarding two different application scenarios.

    1. Without manual reset option

    In a scenario without manual reset option, I wonder if I just have to connect the Delay/M_RST input with GND and put R_ext inbetween and I dont have to connect it to Vdd in any way?

    2. With manual reset option

    Is it correct that when we use an SPST switch, the current consumption will rise only during the time, the button is pressed and is given by VDD / R_ext?

    Does anything speak against using an A-Form reed switch?

    best wishes

    Slev1n

  • Slev1n,

    1. Without manual reset option

    In a scenario without manual reset option, I wonder if I just have to connect the Delay/M_RST input with GND and put R_ext inbetween and I dont have to connect it to Vdd in any way?

    Yes that is correct.  The device will generate a power-on-reset at power up and then start the timer sequence afterwards.  (See Figure 9 in the datasheet for waveform description of this operation)

    2. With manual reset option

    Is it correct that when we use an SPST switch, the current consumption will rise only during the time, the button is pressed and is given by VDD / R_ext?

    Yes, correct.  However, any time RSTn goes low, there will be an additional current path through the pullup resistor.  So for manual reset the increase in supply current will be ~VDD / R_ext + VDD / R_p .  R_p is recommended to be 100 k.

    Does anything speak against using an A-Form reed switch?

    I have not found anything recommending against use of a reed switch for a manual reset switch.  In fact, the device features an internal 20ms debounce circuit, so as long as the contacts of the reed switch stay closed without bouncing for greater than 20ms, a valid reset will be generated.  This feature will also provide sufficient debounce for a standard SPST swtich without the need for additional external circuitry.


    Best Regards,


    David