Powering NodeMCU V3 - what I do wrong?

[XnIcRaM]

I have trouble with powering NodeMCU V3. I found here and not only there than max voltage on Vin pin is 20V. I connect Li-Ion 14,4V accu as Power Supply for my NodeMCU board, but when I connect + from acc to Vin pin then one of diode exploded immediately. I post here photo of my NodeMCU board, diode which was exploded is SMD 1N5819. Voltage regulator on my board is AMS1117-3.3. Now board work but only when I power it up through USB or directly to 3.3V pin (ofcourse with 3.3V) because board work on USB power only then I think that AMS1117-3.3 isn't broken, only diode has beed burned, but the question is WHY?

BTW: my Li-Ion acc 14.4V when I connect it to nodeMcu board have 16.6V and theoretically it's beed 1V over max AMS1117 input voltage but why diode explode and not the voltage regulator?

[justjohn]

Maybe because the diode (1A) couldn't handle as much current as the AMS1117 (1.5A) so it blew up, keeping the AMS1117 from doing so. With a stronger diode, the AMS1117 may have blown up instead.

[XnIcRaM]

maybe... but nodeMCU doesn't need even 500mA

[justjohn]

It's my guess the excess voltage caused the AMS1117 to conduct and short to ground with as much current as the battery was able to provide which blew up the diode. It had nothing to do with current requirements of the nodeMCU. In other words, if the AMS1117 was connected to just a resistor and LED with the diode in the same place, the same thing would have likely happened. Maybe create such a circuit and test just to find out. Likely the same thing if "only" the diode and AMS1117. I'm more of a programmer than hardware guy though, so maybe someone with more of that knowledge will see this and give a comment.

[MrMark]

"diode which was exploded is SMD 1N1519"

Is this correct?  1N1519 is a 4.7 V Zener diode.  The part is there to prevent back feeding VddUSB with Vdd5V, but if you put 14.4V on Vdd5V and the USB is connected, you'll exceed the Zener voltage with no current limiting and it will fail.

The photo shows 1N5819 which is a Schottky diode and would make more sense in this application.

Perhaps your board was built with the incorrect part?  I have a Lolin NodeMCU V3 board at hand and it has "S4" Schottky diodes in that position.

[justjohn]

"diode which was exploded is SMD 1N1519"

I saw 1N5819 (schottky) on his drawing, but you're right that he mentions 1N1519 in his text.

[XnIcRaM]

There is mistake (which I already correct) this diode is ofcourse 1N5819 (schottky)

[MrMark]

Can you tell the orientation of the blown diode?  On my board it is mounted in the opposite direction as the one next to it, that is, the white bar is the side nearest the circuit board edge.  You photo appears to have a white bar on the other side, but it might just be the lighting.

[XnIcRaM]

red arrows show bars on diodes

[mmaxus]

I am dealing with same problem. Even if it is tolerant to as much as 20V, i fed the esp8266 with 12V  then it burned out immediately with intense smoke in seconds. What might be the cause ?

[alwashe]

There is a Problem i asume. I have powerd my NodeMCUs throu 12V and it works for about 24h.

I had a V1 AND V3 wich blew up.

I were not there when they blew up.
They were in closed cases. So they get pretty warm. I don't know.

[RudiAhlers]

I just experienced the same issue. My NodeMCU had 7.2V (accidentally) and it just died. I cannot see any blown components, but heard a soft "puff" sound. So can these boards really not handle anything more than 5V? Even though the board says max 10V.

[rexlee]

I know this is an old post, but I came across this problem yesterday, so the problem is still current.
I have been developing a project using a USB cable and yesterday, it was time to deploy the controller in it's final location.

I had checked
- the current carrying capacity of the schotky diode.    800 mA
- the input voltage of the AMS1117 voltage regulator         40V.
My assumption is/was that the 10uF decoupling capacitor has sufficient rating to handle say >16 V.

Anyway, I turned it on & immediately I saw smoke coming from the input diode (idiot diode) and the 12/14V connecting wire was un-soldered off the pin due to the created heat.  Ie. it acted as a fuse.  However the diode was now short circuited.

I then plugged the controlled back to my lap top via USB and this time the 10uF. decoupling capacitor lit up, so I quickly disconnected the controller.

This morning I removed the faulty diode & capacitor and I find that the controller works 100% OK.

So, my suspicion is that the voltage rating of the de-coupling capacitor is insufficient.  I have no way of knowing what it is, but it might only be 6 Volt or so, which is typical for hi K ceramic capacitors.

With 14V supplied to Vin. this capacitor fails, becomes a short circuit and then causes the idiot diode to fail.

Solution.   Replace this capacitor with a 10uF.   capacitor with a voltage rating of at least 20V.

For what it is worth!