I've ruined two Nano's doing nothing wrong as far as I can tell. I'm powering the Nano and other circuitry from an external 12 volt power supply, via the Vin pin. The AMS1117 (LM1117) LDO is rated for a maximum input voltage of 20 and operating voltage of 15, and I know I'm not exceeding that. USB power was already applied and this configuration worked previously for many power cycles, but this time when I connected the +12 volts to Vin, the Nano responded with smoke and the LDO package is damaged. Evidently the LDO failed shorted (bad!) as all LEDs on the Nano light up and there's no sign of life, so I'm guessing it got a shot of 12 volts when the regulator failed.
There's a diode to prevent interaction between external and USB power so that's not it. I saw no reason for adding an external voltage regulator since there is one on the Nano already, but I think now I will do so and just ignore the onboard regulator.
How 'evidently' is 'evidently'? Are we to assume you know what the reg looks like and you literally see it popped? Are we to assume you didn't short anything or connect anything wrong? What response are you looking for if you're certain it wasn't your fault?
I never rule out "pilot error" but in this case all I did was plug in the 12 volt power lead. By "evidently" I meant that it self-evident - a bubble on the top surface of the regulator where it partially depackaged itself. I removed the LDO and tried the Nano on USB power, but all 3 LEDS still come on and it's not recognized on USB.
I think "lesson learned" - no more than 9 volts to Vin from now on.
Just wanted to get some other opinions in case I was overlooking the obvious, thanks.
Of course the 12 volt return was connected to the GND pin. The LDO which is mounted on the "under side" of the PCB is a 3-terminal AMS-1117 5.0 in SOT-223 package. It's used interchangeably with the LM-1117 but AMS rates the input maximum voltage at 15 vs 20, so evidently that is a critical as perhaps 12 volts with a transient spike occuring on connection is enough to pop it.
I powered up the first board which does not show external damage to the LDO, but when 12 volts is applied to Vin, I also measure 12 volts on the 5 volt pin, which means bye-bye Nano.
I never use on-board regulators with any of my Nanos or WiFi modules. TO-220 three-terminal regulators such as 7805's and 1117's are quite bullet-proof and are as cheap as chips. TO-220 heat sinks are cheap. LM317's can handle more current, handle shorts very well and can have their output voltage adjusted. And, if you do happen to blow up a TO-220 regulator, it is much easier and cheaper to replace than a whole MCU.
Your saying you had the usb hooked up to if so that could be a problem. It's better to hook up the power without the the usb after programming lot's of stuff can go wrong if the USB is hooked up and you then power the chip too.
There have been occasional suggestions that some of the Nano clones may have been made with incorrect regulator wiring and/or sub-par ("counterfeit") regulator chips. So many people just operate off of USB or +5V power that they don't notice, and it's hard to pin down because there are so many "random" manufacturers selling through equally random sellers...
I've never run a Nano clone using the onboard reg, oddly enough. Always used a separate buck converter, usually from a 12V source. I did burn out one diode (the one opposite side the usb port) when some wires shorted. Replaced the diode, all was fine.
Guess I have a reason to keep on with the separate bucks to give 5v to the 5v.
Watch when powering the uC from 5V from an externally powered peripheral as the peripheral external power needs to be on when the uC is connected to USB for programing or you will end up powering the peripheral from the uC 6V regulator instead. I think I lost several nanos that way.
The AMS1117 (LM1117) LDO is rated for a maximum input voltage of 20 and operating voltage of 15, and I know I'm not exceeding that.
Where did you see the "20V" spec? The [url=http://data sheet from AMS I'm looking at[/url] says "absolute max" of 15V, and an un-regulated 12V supply could easily exceed that...
Good point made about an unregulated 12V supply... their output voltage is typically significantly higher than 12V and rely on the load to bring the output voltage down. A Nano isn't much of a load.
I had an odd problem with a Nano... using the onboard 5V regulator to power a few other chips and when plugging in the USB (without an external voltage on Vin), the power light started to dim and flicker, then could no longer upload code. Measurements of the regulator determined it was fried with no output. Was able to replace the regulator and it works fine now, but have since added a slide switch to disconnect the 5V regulator output at the pin whenever uploading code and have had no regulator issues.