This seems like an obvious question that I can't seem to find a good answer for, so please excuse me if I'm just not seeing it. I'm developing some high-speed logic circuits to use on a project with an Arduino (probably a Mega 2560 R3, but maybe a Nano). I expect the external circuitry (logic chips, display screens, and sensors) to exceed the current capability of the Arduino boards. I want to power the entire project with a single AC/DC barrel connect (7-12V) source. My selected power supply will happily deliver 2A@5-12V, so it is more than enough.
I've seen several people suggest that I should just regulate the power supply to 5V and power everything. The problem with this is that powering the Arduino requires 7-12V (Vin), NOT 5V (+5V). I've already fried one Nano learning this before going back and confirming this with the documentation. The +5V pin on the Nano (for example) is the output from its internal regulator. It is NOT intended as a source from a 5V regulated supply. The protection diode switches between the V(in) and USB power when the V(in) exceeds some voltage north of 6V. That protection diode does not protect the Arduino from voltage delivered to the 5V bus from an external source.
So, the best solution I can think of presently (and I am soliciting a better solution) is to draw a parallel line from the barrel connector to power an external 7805 regulator (/w heatsink) and use that supply to power the external circuitry. If I do this and tie the grounds from the Arduino to the ground of the 7805 (short distance between them), I think that should suffice. But it just seems like a "messy" solution. I would appreciate any advice (other than "go ahead and ignore the specs and use the 5V bus on the Arduino to power everything).
I'm surprised that at least for a Nano you can't power it from an external regulated 5V supply to its 5V pin. I could swear I've done this many times. There may be a small current that flows through the regulator to ground, but not enough to cause any problems. And the Nano has a diode in the USB power line, so current can't pass from the 5V pin back into USB. In fact, USB is applying power to the 5V pin through that diode. So applying power at the output of the Nano's regulator seems to work. Can you describe the setup you were using when you fried the Nano? Do you still have the fried Nano?
Well, you'll probably have to do that anyway. The regulator on the Nano is small and not heatsinked, so it can only provide so much current before it overheats. So I guess you're saying you would have a higher supply voltage feeding at least 7V to the Nano's Vin pin, and the 7805 supplying everything else, with all the grounds connected. That should work assuming all the other parts work on a 5V supply. But the question remains why you can't supply everything from a single 5V supply, including the Nano at its 5V pin.
The voltage regulator (not heat sinked) on a genuine Mega can dissipate about 1.2 Watts before overheating and shutting down or burning out. The maximum current = 1.2 / (voltage in -5V) so, with 7V in, about 600mA, with 12V in, about 170 mA.
Many of the cheap Nano clones can't take full 12V on the VIN pin with minimal load without blowing the regulator.
It is a variable voltage DC power supply. I’ll probably get a standard single voltage supply for the final project. As to which Nano, it was the standard basic Nano. None of the 33 or ESP or any of those.
It was a standard Nano. It did operate with the external source providing 5V to the bus. But when I plugged in the USB at the same time, something happened. The board no longer will allow me to power it via the usb port. Not just when the external power supply is connected. If I try to power the board with just the usb, it is dead. It will still work with external power and the usb will still work for data. I believe that the protection built in does not trigger until the external voltage on the VIN pin exceeds about 6.5V. I don’t think that the protection was ever intended for the 5V bus. The spec sheet is very clear… external power can be from a regulated 7-12V source onto the VIN pin. I don’t see any official reference saying that external power can be from a regulated 5V on the 5V pin (and this would be a VERY convenient option; as many projects need more current than the Arduino can provide).
As the schematic shows, USB power flows through the diode to the 5V pin. And from your description of what still works, it appears the diode has blown. I don't know what could cause that from having both supplies connected at the same time. The whole point of the diode is to allow for that. Maybe something in the external supply was sinking large amounts of current for some reason. But I don't see what would cause that. Maybe @jim-p, who is an actual engineer, can offer an explanation.
I have something called a breadboard power supply, which can provide either 3.3V or 5V to the rails of a breadboard. I have often powered a Nano by connecting its 5V pin to the 5V rail. So I just don't think that's a problem. But maybe connecting that and USB at the same time is a problem, although I don't see why that would be the case.
The circuitry in the UNO R3 (and the MEGA 2560) allows to feed current back to your PC when powered trough the 5V pin. The Nano (classic) uses a diode to prevent that.
Boards I’m familiar with allow power into the barrel jack or VIN pin while connected to the PC (in case of UNO R3 it needs to be over 6.6V). This comes with a severe limit on the max current, as described in post #4.
When powered by USB the UNO R3 allows you to pull current from the 5V pin with a limit of 500mA from your PC, which is more than when powered from the barrel jack or VIN. Thus knowledge about the board you’re using is essential.
Because all these conditions and considerations a general advise is given not to use the 5V pin.
Thank you for the schematic. This is actually a name brand Nano - not a knock-off. If that is an actual schematic (as opposed to a functional schematic), then I cannot explain why I would have had this issue. It also leaves me quite confused as to why the official documentation does not suggest that powering the Nano from a direct 5V source is acceptable. After all, it is not at all odd to have a project that can exceed the power requirements offered by the Nano board. Perhaps I just got a bad Nano. I have 3 other Nano boards that I can try to duplicate the issue with. Thinking that I learned my lesson NOT to power the board with input to the 5V pin, now I'm considering it again to see if it was a 1-off issue.
It discusses how it might have been fried, but basically, there had to be a short to ground at the 5V pin somehow for the diode to go bad. If you have a multimeter, you can test the diode.
So your fried Nano still works if powered at Vin. Does it also still work if powered with 5V applied to the 5V pin? If it does, then the Nano proper is still good, and the diode is what's wrong.
Based on your post (and the posts of others on archived forum posts about similar issues), I'm going to try again with a different Nano. Right now, the only thing I can think of is that I may have been operating my computer without the charger plugged in and therefore at the time I connected the DC adapter to the board, the two power sources may have had different ground potentials and an electrostatic discharge fried the diode as I was plugging the second power source to the circuit. That would surprise me however; as these adapters are designed to connect ground before power.
Or, the external supply may sink lots of current when it's powered down. It could be a lot of capacitance on its output that if discharged could initially appear as a dead short to the USB supply as the capacitance charges up. And these days USB can supply a lot more current than the original 500mA limit on USB2. I would think the diode would survive 500mA, but I don't know how much more than that. Of course I'm just guessing here, but it could be something like that. Then you would power up the external supply before connecting USB.
In any case, your fried Nano may still be perfectly usable except for being powered from USB. And if you replace the diode, it would be like new. That's assuming the diode is the problem, which you should check with your multimeter.