I am using a cheap chinese ProMini168 for a simple USB-to-RF22 bridge: The respective device is permanently plugged into some USB port of my PC and I am using it to communicate with other distributed devices from some Java based application program. The device is powered directly from the USB 5V supply.
The device worked nicely but after about 2 month of uninterrupted uptime, the device was "dead" (display was blank and PC application did no longer detect it's presence) when I checked it today (I had not actively used it for some days and I don't know the exact time of death). Turning the device off/on did not help.
Finally I just re-flashed the program on the ProMini168 - which immediately fixed the problem, i.e. for some reason the Flash memory on the ProMini168 had apparently gotten corrupted during normal operation.
Are there any known issues that might explain this problem? (I want a robust device that I do not have to reflash every once in a while - and I'd like to improve my device accordingly.)
PS: The main grid power supply here is of notoriously bad quality, i.e. during thunderstorms mains voltage may fluctuate between 200 and 250V and in the past I have already lost commercial electric equipment to this effect... however I am not aware of any thunderstorm that might have struck yesterday (or in the past weeks) so that may or may not have played a role here.
PPS: Some days ago extended use of the 3d-accellerator in my PC had overheated the complete PC causing it to shutdown (I do not know if that might have had any effect on the USB 5V supply..)
I think I just found the problem: Win10 seens to dynamically assign the names to the different USB ports - somehow related to the order in which devices are plugged-in. It seems that the 1st USB-to-TTL adapter that I plug-in on my PC always becomes "COM5" the next adapter then becomes "COM4" - and it makes no difference on which of my various physical USB sockets I plug-in the stuff. If stuff is plugged-in in a different sequence then adapters are "suddenly" connected to different "COM" ports..
The USB-to-TTL adapter that I am using within my device is the same kind of component that I am using to program ProMini's via the "default" upload protocol. If I remember correctly (it's been some time) I had originally tried to setup the device such that it would not work to accidentially reprogram it using that adapter and I am actually reprogramming its ProMini using a separate programmer (unfortunately it seems that attempt has failed).
When I recently tried to re-program a different ProMini (within the IDE the COM* port is typically fixed for a specific project), the assigned "COM" names must have gotten mixed up and I actually re-programmed the wrong ProMini by mistake (a mistake that slipped my attention since I did not immediately use the "newly programmed" ProMini).