In March, I purchased a 3-pack of Nano Every's. The first one, I uploaded several sketches while testing/debugging a project; when I was finally ready to upload the final version, the board heated up and burned out during the upload - no visible damage, but I could smell the parts burning; and there were lots of error messages from the IDE. After I disconnected it and let it cool off, I reconnected it to the USB port and got nothing...no power LED, and no recognition by Windows. It is completely dead. :-X
I setup the second board, uploaded the sketch, and no problems...it is still running today. ;D
The third board I've being using for testing several projects, until today. During an upload, it too got very hot and the IDE displayed lots of error messages. I disconnected it and let it cool off. If I connect 5 volts to it. the power LED comes on, and the whole board gets very hot...obviously no longer usable. :-X
The two boards that burned up had had sketches up loaded to them several times before they died;
The board that is still running had been programmed only ONCE.
Also, note that at the time of the burnings, they were only connected by a USB cable to my computer, nothing else.
Is there a limit to how many times a sketch can be uploaded to these boards; or is there a bug in the uploader software that creeps in once in awhile?
Has anybody else had a Nano (or any Arduino) fry while uploading?
I can't accept a 66% failure rate on these boards; if this is normal, then I'll have to consider other boards.
I have burned 2 Arduino Nano Everys.
I can confirm it has to do with uploading because the one I just burned had otherwise been happily working with USB + Power supply(5V) at the Vin pin simultaneously for many months during a project's testing phase. Also worked with power only or with USB only. Also capable of uploading regardless of the presence of power supply. At some seemingly random time, during an upload with power disconnected, it failed and shut down completely (all LEDSs off, USB disconnected, etc).
My first Nano Every died very early and come to think of it, it must have been just when I tried to upload something to it for the first time (also attached to external +12V power). So this whole thing has to be due to the upload procedure, among other factors.
Aside from the inability to upload sketches, both boards in their entirety got very hot and smelled burned. The first one had smoke come out some component of it (could not visually clarify). The second had smoke come out of the schottky diode next to the USB. So it must be the Schottky diode that got trashed in both arduinos.
Very surprisingly, the Every I just "burned" actually managed to get programmed when I isolated it from everything and powered it with USB only. So the atmega4809 on it is not toast. At least not completely. I am not going to connect it to anything for now though, in order to prevent any further damage. Maybe my first Arduino is salvageable too.
I am going to swap the schottky with a new one and see if the problem gets fixed.
Wait till you see what you have to go through to get a replacement under warranty; it isn't worth it.
nah...I'll replace the schottky myself with a similar one. Difficult weird hackjob but I don't have much to lose at this point. Especially my first Every seems dead already anyway.
Fortunately the diode is the component with the biggest pads in there.
I'm 64 years old with less than steady hands; I wouldn't be able to replace something that small, even if I could identify it.
I would like to know what results you get after you replace the diode on yours.
When I ordered 3 more Everys, one was dead on arrival. I think the manufacturer has a design or manufacturing defect with these boards; I think more people will be complaining about the failure rate.
stecoop:
I think the manufacturer has a design or manufacturing defect with these boards;
Its possible they do. I have seen people report fried boards happening in similar ways to ours and also people talking about Nano bundles having 1 or 2 out of the three break very early. Granted, I should have protected my own Every better but other Arduino boards have taken worse beating and ended up just fine.
The Every is a nice little board for a good price if you want to avoid the chinese Nano clones with the knock-off FTDIs or the questionable CH340s. But if they keep frying on me no matter the ever increasing protections I take, I 'll have to abandon them. We 'll see.
stecoop:
I would like to know what results you get after you replace the diode on yours.
Yes, I will let you guys know. I had started a thread myself around here about a burned Nano Every, so if I have any progress, I 'll link to that thread and post.
stecoop:
I wouldn't be able to replace something that small, even if I could identify it.
Its a finicky part, indeed. I only have a soldering iron and that makes it even more difficult since I am mostly a noob. The elegant procedure would require hot air, solder flux, desoldering braid, tweesers and of course, a suitable schottky diode replacement.
I already managed to replace the older Every's with a normal diode (didn't have a schottky yet) but it still doesn't work. I'm not surprised, that board was pretty dead anyway. Also, the diode I used was not a correct replacement.
Anyway, I 'll see if I can do anything in the folowing weeks with the newer Every that is not completely dead. That one has a burned diode but also a short somewhere else but still managed to communicate, just after the diode started smoking.
One of the Nano Every's I ordered was "dead" on arrival - when connected by a USB cable to my computer, the computer would NOT see it, only the green power LED on the board would light up, barely.
Tonight, I tried connecting 5 volts to it through the Vin and Gnd solder terminals, but no USB connection; the green power LED lite up, the "Blink" program began running, and the yellow LED began blinking. I tried connecting it to the USB port WITHOUT the external 5 volt supply and got the same results I started with - dead.
After playing with it for a little while, I found that if I FIRST connect the external 5 volt supply, then connect the USB, IN THAT ORDER ONLY, it will work; I was even able to upload a modified version of the blink program (different blink rates) to verify it was working. It did get a little warm during the upload, but not unacceptably so, and not "flaming" hot like the other two that burned up during their uploads.
Clearly, what ever the defect is, it is related to that USB port and getting power through it.
The spec says Vin needs 6-21V. I have also been powering mine with 5V and now due to the link below I think it might have contributed to the fault in the long run:
Someone in there says that if Vin is lower than 5V, which could momentarily happen in real life,
the output mosfet acts as a dead short between Vout and ground
I don't know if they are correct but it sounds like a possibility, at least for my burned Nano Every.
stecoop:
One of the Nano Every's I ordered was "dead" on arrival - when connected by a USB cable to my computer, the computer would NOT see it, only the green power LED on the board would light up, barely.
Disappointing news. I will use up (since they seem to have an "expiration date") the rest of my Nano Everys and move to something else. At the same time I 'll try to protect them better by powering them with more than 5V at Vin and with extra diodes and PTCs, and we 'll see.
I did a small investigation on my Nano Every. Its quite a big write-up so I haven't finished it yet. I 'll upload it when ready. Spoiler: I think the MPM3610 voltage regulator got shorted internally (5V with GND) and that also fried the onboard Schottky.
burner_:
The spec says Vin needs 6-21V. ....
That's interesting because I have 3 Nano Every's (now 4, actually) running with just 5 vdc on the Vin pin; and they're running just fine.
I looked at the "article" you linked to; it seems like there are a lot of inconsistencies in the behaviour of the USB port and Vin pin. Me thinks there is a serious DESIGN flaw in the power systems for these Nano's.
stecoop:
That's interesting because I have 3 Nano Every's (now 4, actually) running with just 5 vdc on the Vin pin; and they're running just fine.
Very good to hear.
stecoop:
Me thinks there is a serious DESIGN flaw in the power systems for these Nano's.
Yes, I am thinking that too. I 'd say there needed to be another schottky that would prevent current from flowing OUT of the Vin pin and to whatever it could be connected to in a real life PCB. Who knows what else...
(Edit to add: When powered by USB, the VBUS electricity goes directly to the internal +5V bus of the board and the MPM3610 regulator is not used. For some reason, however, the regulator sees the 5V on the +5V bus and OUTPUTS 5V from its Vin pin, from which it would normally draw current instead of outputting. Essentially, the arduino when powered with USB only is trying to power up everything and anything else connected to the board's Vin pin which would normally be served by a beefy power supply. If the rest of the peripherals on the rail are hungry, the USB overloads, the Schottky diode burns and/or the regulator fries and shorts the GND and +5V pins internally. They should have put another diode to prohibit this weird Vin current backflow. I will put one myself next time.)