Was testing some signal filtering on test board and forgot that it was fed by Arduino Uno SMD R2. Connected a 1000uF cap between +5V and GND after which Uno freezed.
Not able to upload anything on the board anymore. It's recognized by Windows, +5V & +3.3V are there, RX LED blinks few times during download but then it ends up with avrdude: stk500_getsync(): not in sync: resp=0x00 error.
Green status LED is constantly on. Never blinks. If I supply the board with power jack the regulator gets fairly hot. Still, don't really understand why would the high current draw from +5V supply had broken the Atmel chip itself.
Any ideas which components I should check or is it an obvious game over?
The 5V circuit can be traced using the schematics on the UNO products page. It might help you see what could be affected by the cap on the pin you tied it to.
Thanks guys. A bit hard to remove 328P as, like said, it's the SMD version. I know it took quite a huge current rush with that cap but still wonder what could have broke down if/when the regulator seems to be ok and that current should not flow through the uC. Also tried to supply the board from other feeding points without luck. Have to try to connect it via ISP soon as I can get a programmer/another duino.
I really need to put my brain into gear before I put my responses into motion! Sorry. Simply was thinking UNO even thought I did read SMT... my bad.
In theory, an inrush of current from the power supply because of a high-value capacitor should not have taken out the uC... provided the current stayed solely on the power rails. This is like shorting out the battery on your automobile - one would not expect the radio to "blow" because of that as all the current went from Positive to Negative in the short. As mentioned earlier, referring to a schematic of the board and identifying where you placed the capacitor (momentary short) may reveal what devices were damaged: regulator and re-settable fuse perhaps... in theory, the fuse should open until the overload is removed, but in the case of a capacitor, the fuse may never open due to the short time of the high current.
Reaching for a cause, I might suggest that the capacitor was charged with a voltage greater than +5 volts when connected? This scenario would have taken out the uC if the voltage on the capacitor was greater than something like 5.6V. Large capacitors can hold a charge for some period of time, so if you were using the cap ahead of the regulator and then moved it to after the regulator, the uC is likely zapped.
The cap was plugged directly between +5 and GND on test board (wired from arduino to the test board). Cap couldn't have had more than +5V as it was there already earlier so overvoltage is out. Any change that 328P might corrupt it's memory during undervoltage?