CCS/TIDA-00489: TIDA board failures after automatic soldering and molding

Genius 5170 points

Replies: 4

Views: 49

Part Number: TIDA-00489

Tool/software: Code Composer Studio

Hey guys,

we use the tida 00489 board in our sensor design. We noticed, that some of our sensors stop working after soldering in the battery (we use a different battery type) and molding the board with PU mold.

Regarding the manufacturing process, we flash the board and perform a functioning test, during the test the board is powered externally via the JTAG connection. Afterwards, the battery is soldered onto the board, placed into a housing and molded.

We noticed, that after this process some of the sensors did not work anymore, whereas the others worked as expected. After breaking of the housing and manually resetting the sensor, it worked fine again. Since we do not have the possibility to restart the sensor (we do not have an on/off button) the sensor can only be restarted by irreversibly breaking up the system. Thus we were wondering if any experience reports are available or if you have any idea how to handle this. Since the device is completely encapsulated, we cannot debug it since we cannot attach anything and the wireless transmission is, as mentioned before, not working anymore.

Could there be a misfunctioning of the RF module due to overheating during soldering?

We somehow think that either the soldering or the encapsulating is the reason, though the latter one is performed at rather lower temperatures around room temp.

Hope you guys can help us out once more. Dont hesitate if you need further info.

Best wishes

Slev1n

4 Replies

  • Hello,

    The engineer who designed this board is out of the office at the moment but hopefully can weigh in once he is back later this week.  In the meantime, though, I did have a few questions/comments:

     - Is the firmware published for the reference design used directly, or has it been modified in some way?

     - To reset the system, are you using the nRESET control or disconnecting the battery?  (Or, do either work to restore functionality?)

     - Is any testing performed after the battery is installed but prior to encapsulation?  If so, is this issue ever identified at that stage?  I ask this because if there were an issue resulting from encapsulation (like overheating or an unintentionally made/lost connect) then I wouldn't generally expect it to be resolved via a reset.

    Regards,
    Max

  • In reply to Max Robertson:

    Hi Slev1n,

    As Max mentioned in the previous post, a test after the battery is soldered would be a good idea to let you know what part of the manufacturing process is causing the issue.

    There are a couple of previous threads we have discussed in the past which I've linked below.  Have the previous recommendations on the encapsulation been followed?  Has the watchdog proposed in the second thread been implemented or could it be used here?  Is the product being referenced in this thread related to these previous two?

    https://e2e.ti.com/support/tools/sim-hw-system-design/f/234/p/817954/3028753

    https://e2e.ti.com/support/tools/sim-hw-system-design/f/234/p/934060/3464918

    Regards,

    David

  • In reply to Max Robertson:

    Hey Max,

    thank you for taking your time.

    1.

    Max Robertson
     - Is the firmware published for the reference design used directly, or has it been modified in some way?

    We have made several changes in the software and we have of course checked if there might be any bug. But when we flash sensors with the same software which have already a battery attached, we usually dont have similar issues. Though we do not encapsulate these types of sensors, since they are only for indoor testing.

    2.

    Max Robertson
    To reset the system, are you using the nRESET control or disconnecting the battery?  (Or, do either work to restore functionality?)

    We have modified the layout a little bit. We kicked out the RESET button and we have to solder in the batteries, thus the only way to restart the device is to short-circuit it for a short moment. This is surely no good solution, but the only way and quite reliable, since removing the encapsulation material is very difficult.

    3.

    Max Robertson
    Is any testing performed after the battery is installed but prior to encapsulation?  If so, is this issue ever identified at that stage?  I ask this because if there were an issue resulting from encapsulation (like overheating or an unintentionally made/lost connect) then I wouldn't generally expect it to be resolved via a reset.

    At the moment there is NO testing between these two steps but it is surely necessary to determine whether its the soldering or the encapsulation. However, I rather suspect the soldering step, either due to temporary overheating causing the device to go into an undefined state or due to repeatedly restarting during battery soldering.

    best wishes

    Slev1n

  • In reply to Slev1n:

    Hey David,

    I just noticed I have overseen your post. And you are correct, that we are still talking about the same device, though now the devices stop working after manufacturing and not within the field.

    1. Regarding the watchdog, currently its rather economically to take the risk of some sensors to stop work compared to feature every device with a watchdog, since we were able to determine the main error source, which was still water/humidity causing a short cut and cold solder joint.

    2. We have a slight worse RF performance but still a long enough range. Regarding a protection of the chip during encapsulation, we could cover it with silicone first, which might take some of the mechanic stress from the chip housing. Regarding the sensor, with an earlier type of encapsulation we once damage the PIR, but with the new material, the sensors seem to work properly.

    We are going to encapsulate a new badge soon, implementing a new testing step after battery soldering. If I have new info, I will report back.

    best wishes

    Slev1n