Supercaps... 100F is always startling to me.
I cannot go into too much detail as there are IP concerns at stake. This design failure concerns a device for the purpose of....hmm...controlling a sight where customers might come to fill there "something", with uhhhh, a few gallons of "something." Anyway, the predecessors and contemporaries of this device use a rechargeable battery pack to ride out power failures and brown outs. The current, just released, not fully tested version uses supercaps!! in place of rechargeable batteries. The microcontroller on board needs about 60 seconds of power so that it could perform an orderly shutdown, store transactional data, etc.. Everything squared away and it it's place you see.
Failure: The supercapacitors were laid out on the board so that they are permanently in parallel with the microcontroller power bus. Because the supercap never lets the micro fully power down, the micro could never actually power-cycle and restart it's clock. The result is an unresponsive "brick" and from the customer's perspective, a defective/malfunctioning/broken "call it whatever you want, it's not working!" unit. Well, it could power-cycle if you physically unplugged the device from it's 85 to 265 VAC source and wait for the modest current draw from the system to completely drain the charge from those supercapacitors. Unplugging an actual installed unit is not a practical option because units in the field get there power from a conduit.
Also, the unit was supposed to be able to perform it's own resets. Besides, periodically having customers unplug their units for 20 minutes at a time is not a solution. The fix for this? I don't know and it's not my job. I just get to hear about the failure from other engineers.
Have any of your engineering projects ever failed due to factors beyond your control? Share your story here!
See more news at: