Fried sensor board with external AC/DC adapter

[mardah]

I have fried several sensor boards when using an external AC/DC adapter supplying 12V/1A. The voltage is regulated with a MCP1703 with 1u caps on both input and output.

I think the problem appears when I unplug and plug in the DC jack to quick? Not sure though. Sometimes just the voltage regulator fries and sometimes the ATMega
fries as well.

If it doesn't fry the first time I plug in the DC jack it can run for a long time.

What could be the problem? Build up of voltage in the AC/DC adapter, the caps? Voltage spikes?

When running on two CR123 in series I have not seen any problems with the sensor board.

[mardah]

[sundberg84]

A read somewhere here that cheap clones use cheap/fake voltage converter and supplying 12v to RAW is living on the edge even if 12v is within limits.

Maybe you have a bad batch of voltage regulators?
Did you try with another AC/DC converter? It can limit your search.

[mardah]

The MCP1703 does provide 3.3V until it fries. But yes, it could be a non genuine voltage regulator.

[sundberg84]

How about the HLK-PM01 or use a AC to 5v DC converter (Iphone plug)?
Then you dont need regulate your voltage twise.

[Yveaux]

@mardah what's the maximum input voltage for the MCP1703?
I've fried some AM1117's supplying '12V' from an unregulated adapter.

[mardah]

The MCP1703 should handle up to 20V input. Maybe the input cap is 10V , I guess it would handle 12V most of the time?

[Yveaux]

@mardah said:

he MCP1703 should handle up to 20V input.

Datasheet states 18V as absolute maximum (http://ww1.microchip.com/downloads/en/DeviceDoc/22049e.pdf).
Also check the maximum on the cap, as you already suggested.

[m26872]

I've fried a lot of Arduino boards by powering on my breadboard AC/DC supply while still having the FTDI port supply from Laptop connected or vs. (Laptop is on its own AC/DC.)
Change Atmega has always saved the boards then, but I don't use the v-reg very often.

[Yveaux]

@m26872 ground-loops... Terrible

[mardah]

I soldered just the two caps and an MCP1703 on a new board and have not manage to fry it. So maybe there is another cause. I'm suspecting the WS2812, it's new on my latest batch of sensor boards. But I don't understand how it could destroy the voltage regulator or ATMega.

Maybe something is wrong with the PCB layout causing some unwanted currents?