My circuit blew up from wall adapter?

[WillHuang]

I created a small breakout module that would boost a li-po battery to 5V. I've also included a battery charger IC in my module that would, not only charge the battery through a USB, but also provide 5V out. I've attached my schematics down below.

Now, if I were to plug in lipo battery and the USB using my laptop's usb power port, or the usb power port from my surface pro's charger, the circuit works fine. I see my lipo's voltage steadily moving up. However, if I were to swap the USB power to a wall adapter (I have a picture of the one I used attached below), my board got very hot and would not produce a constant 5V out anymore - it's giving out 2V instead. It's also very likely that the switching regulator IC got burned up (I know this because it happened once before, and had to replace the IC).

I'm not too sure what's happening and would like to get some insight from you experts! 

[amyk]

Do you have a scope? Look at the voltage on the output of the adapter if you do. It probably has a lot of high-frequency ripple, and if you do not have enough decoupling in your circuit, that can cause problems.

[amyk]

RPi may have better decoupling...

[Alex Trofimov]

Hi!

If there's something horribly wrong with your circuit it might have been OK while you used laptop port because of current limiting. But with 2.5A capability of the wall adapter it blew)
I suppose, you checked adapter's voltage under and without load, I just mention this because you didn't point it.

[helius]

As shown in the Linear app note, the inductance of the power adapter and cable can cause a voltage spike that blows up the device.
While you should understand the cause of this problem and eliminate it in your designs, it means that when powering some random device you should observe the following procedure:
1. connect the power adapter to the device first;
2. connect the power adapter to the mains last.
This is somewhat counter-intuitive, because in my experience most people think that plugging a device that's already "on" into the mains is "more stressful" (the same way you avoid replacing light bulbs when the switch is on). But they're wrong.

[WillHuang]

Thanks for the replies everyone.

Do you have a scope? Look at the voltage on the output of the adapter if you do. It probably has a lot of high-frequency ripple, and if you do not have enough decoupling in your circuit, that can cause problems.

I looked at the scope and I don't see any ripple voltage on the output of the USB

Ceramic Input Capacitors Can Cause Overvoltage Transients

http://cds.linear.com/docs/en/application-note/an88f.pdf

See this thread
https://www.eevblog.com/forum/beginners/transient-positive-voltage-spikes-killing-mosfet-drivers/25/

one of many, wondering why when they use a power brick all of a sudden their circuit doesn't work. That application note should be a sticky or something.

Thanks for the app note- it was a good read, and it could be the problem? However, if I remembered correctly, the circuit was working fine with just the wall adapter alone - I need to double check on this when I solder new parts in. I didn't get a chance to measure transient voltages on the board yet either.

Hi!

If there's something horribly wrong with your circuit it might have been OK while you used laptop port because of current limiting. But with 2.5A capability of the wall adapter it blew)
I suppose, you checked adapter's voltage under and without load, I just mention this because you didn't point it.

This was one of the first things I thought of when the board blew up, but it's also something that's confusing me quite a bit too because there was no output load when I powered on the module.

I did a little experiment today.

1) I measured the output voltage of the wall adapter: it was read about 5.25V on my 6.5 digit Agilent multimeter.
    Question: if the boost IC receives more than 5V in, but is suppose to generate 5V out, how would boost circuits normally behave?

2) I had to remove the boost IC again as it failed short. After removing the IC, I tested the board with the Li-Po battery/wall adapter, and li-po battery/laptop port again. Here's what I observed:
       - For Li-po battery/laptop usb port: the battery voltage started off at about 3.8V and steadily increased from there. I ran it for a couple of minutes and no problems occured. I thought that this was normal
       - For the li-po battery/wall adapter: the battery voltage started off at about 4.15V?! and it increased from there. I didn't get to see if it would reach all the way to 4.2V and stop charging because I noticed after a minute or so, the board got very hot. I also measured the input voltage of the USB and it's at about 4.9V or so. I'm very confused here as
          a) How come the battery would start off at a higher voltage this time?
          b) The fact that the board got very hot, and the wall adapter voltage dropped from 5.25V to 4.9V, suggests that somewhere in the circuit it is drawing more than 450 mA (450 mA is what I configured the battery charger IC to draw).