Short circuit broke my Arduino Mega?

Hi!

I am trying to build a 3D printer, and for that I have used the Arduino Mega 2560 in combination with the RAMPS 1.4.

While testing a motor on this configuration, I somehow got a current spike, which melted the insulation of a breadboard wire, got through to my laptop, and to the ground. The laptop switched off, and after that the Arduino Mega would not communicate anymore with my PC (tried on different PCs).

I checked the voltage regulator on the Arduino (voltage seems OK), and the red led is on when I connect it to a PC. However, my PCs cannot find the board, and as such I cannot program it anymore.

Does anybody have an idea how I could rectify this? Or is this board lost to everybody?

Thanks!

Hell, that must have been a HUGE spike to melt wires like that...

Did any Magic Smoke™ escape from the Arduino? If you melted wires, then chances are you melted chips too...

Time to let it go....

I'll repair it for you... It'll cost you about the same as buying four new Megas.

Hey! Maybe some extra info about what actually happened. It might some time help another poor soul :slight_smile:

In attachment you find a small schematic overview of the setup.

To program my arduino, I connect it through USB to my laptop. Every time I wanted to do something with the board (eg replace the motor driver), I would disconnect the power from the board. I use a 12V, 5A power supply, which was connected through a breadboard and a switch to the RAMPS 1.4.

I switched on and off the supply to the RAMPS numerous times to change things on the boards, until the moment that a short (?) generated that much current that a jumper burnt through, and my laptop switched off, not to be revived. BTW: the laptop was grounded through its power supply. After inspection, I found that a wire just after the switch had burnt through, but should not have caused a short circuit (was not grounded nor in connection with other conducting surface).

Additionally, I was not able to reprogram the RAMPS 1.4, as the arduino board seemed to have lost USB connection functionality. The board itself seemed to still work.

Hinsight, this might not have been the smartest setup:

  • Small wires on breadboard connecting the high current power supply with the RAMPS & motor (Nema 17).
  • Switching would result in high currents throughout the system
  • The grounding of the laptop was probably better than that of the power supply

My questions:

  • How can I avoid such a scenario from happening again? A normal printer would also be connected to the laptop through USB and does not result in a current spike up to the laptop?..
  • How can I avoid the high currents through the system at startup (if this is the root cause)? Can I further isolate the motors?
  • How to protect the USB interface from any short circuit/high current?
  • I understand that there is no way to repair the Arduino Mega. In your opinion, did the USB interface break?

It’s been quite a scare… My laptop now is at a shop for inspection. I think it’s possible to fix it, but I’ll wait and see.

I hope to be able to avoid disaster as this happening again!

Circuit.pdf (17.9 KB)

I suffered a similar issue while building my 3d Printer (wired up the Endstop sensor board wrong) and created a short.

The fix was easy. I un soldered the LDO 5V regulator and replaced it with a new one. Works fine now.

The way to avoid it is to be super careful. In my case, I had used a bad reference document.

Note: It was just chance that I caught it fast enough that only the regulator fried.

majenko: Hell, that must have been a HUGE spike to melt wires like that...

Did any Magic Smoke™ escape from the Arduino? If you melted wires, then chances are you melted chips too...

Majenko, I was on top of it all when it happened. No smell, no smoke. I inspected the Arduino board, and found nothing out of the ordinary. The only things that were out of line, was the melted wire, the communication that was not working for the Arduino and my laptop that gave its spirit to higher fields...

My concern definitely is less about reviving the Arduino board than avoiding anything like this happening ever again... Any suggestions as to lower chances to get current spikes are greately appreciated! :)

The Ramps 1.4 board is designed to move the motor load far away from the arduino board itself... but it won't protect you from shorting the 5V rail.

pwillard: I suffered a similar issue while building my 3d Printer (wired up the Endstop sensor board wrong) and created a short.

The fix was easy. I un soldered the LDO 5V regulator and replaced it with a new one. Works fine now.

The way to avoid it is to be super careful. In my case, I had used a bad reference document.

Note: It was just chance that I caught it fast enough that only the regulator fried.

Hi! Thanks for the advice! I read about that potential solution somewhere. As I'm more kind-a software guy, I'm not really used to fiddle with electronics, so I'm learning every day :)

I checked the voltage over the LDO 5V regulator, and it gives:

0.52V 4.88V

This is in line with a (working Arduino Uno) which has

0.50V 4.99V

Do you know if a not working LDO would still give this voltage drop?

And indeed, had I not been there, things might have been worse! Makes me wonder which precautions I have to take to not burn my home when I build a 3D printer! Any suggestions for extra layers of protection are welcome!

pwillard:
The Ramps 1.4 board is designed to move the motor load far away from the arduino board itself… but it won’t protect you from shorting the 5V rail.

I imagine you used 5V for your setup? I used 12V which came from my power supply. I’m still not sure how the short circuit could happen, as there were no wires connecting to eachother, or to ground…

Hi, have you tried another USB cable?

Tom........ :)

Well, we will probably never know, as we have absolutely no idea what you had connected to what.

Do try another USB cable, and inspect your Mega board very carefully.

If it works fine except the USB, you might have popped the USB VCC fuse. I'm under the impression that detecting the USB 5V is akin to the on switch in serial communication protocol, without it it assumes something is wrong or not connected.

If you want to know without a doubt that it works you need a programmer, one that connects to the SPI pins. If you can't interface then its dead. You'll need to check both the serial chip and the 2560.

But first see if you can measure the 5V before and after the fuse.

The USB polyfuse is supposed to be self-resetting... but... it's possible that the USB-SERIAL bridge IC did not survive.

pwillard: The USB polyfuse is supposed to be self-resetting... but... it's possible that the USB-SERIAL bridge IC did not survive.

The USB polyfuse is there to protect against short-circuits, not overvoltage.

Hi Please answer Reply #10 and #11.

Tom..... :)

fungus: The USB polyfuse is there to protect against short-circuits, not overvoltage.

Well, the original description was "a current spike", which certainly does sound like one description of a short circuit. :astonished:

Paul__B:

fungus: The USB polyfuse is there to protect against short-circuits, not overvoltage.

Well, the original description was "a current spike", which certainly does sound like one description of a short circuit. :astonished:

There are different types of "short circuit".

  1. Power -> GND - This is where a power rail connects direct to ground and massive currents ensue. Fuses blow.
  2. Power -> Power - This is where two different voltage power rails connect together, and high voltages get injected into low voltage components. Components blow and melt - as a result fuses may also blow if components end up as short circuits themselves.
  3. Cross-component - This is where a component effectively gets removed from a circuit and replaced by a piece of wire. Circuits do strange things, and currents may increase, voltages may increase or decrease depending on the circuit.

When you get an over-voltage situation on semiconductors they very often turn into short circuit as the silicon breaks down. That causes increases in current, which then causes melting of the bond wires, which turns it into an open circuit. So you get a voltage spike, which then causes a current spike, which then causes the release of the Magic Smoke™.

Hi All!

Sorry for the late reply, I was at work and did not have access to my PC...

I try to answer your questions:

TomGeorge: Hi, have you tried another USB cable?

I did try another USB cable (actually a few). I even switched computers. However, to no avail. The board itself looks (from the outside) perfectly fine.

randomname: If it works fine except the USB, you might have popped the USB VCC fuse. I'm under the impression that detecting the USB 5V is akin to the on switch in serial communication protocol, without it it assumes something is wrong or not connected.

If you want to know without a doubt that it works you need a programmer, one that connects to the SPI pins. If you can't interface then its dead. You'll need to check both the serial chip and the 2560.

But first see if you can measure the 5V before and after the fuse.

Probably I won't be able to repair the board in any case, but it's great to learn from ones mistakes! I am not sure what you mean with before and after the fuse. I checked the voltage after the LDO 5V regulator, there it was 4.88V.

majenko: There are different types of "short circuit".

  1. Power -> GND - This is where a power rail connects direct to ground and massive currents ensue. Fuses blow.
  2. Power -> Power - This is where two different voltage power rails connect together, and high voltages get injected into low voltage components. Components blow and melt - as a result fuses may also blow if components end up as short circuits themselves.
  3. Cross-component - This is where a component effectively gets removed from a circuit and replaced by a piece of wire. Circuits do strange things, and currents may increase, voltages may increase or decrease depending on the circuit.

When you get an over-voltage situation on semiconductors they very often turn into short circuit as the silicon breaks down. That causes increases in current, which then causes melting of the bond wires, which turns it into an open circuit. So you get a voltage spike, which then causes a current spike, which then causes the release of the Magic Smoke™.

Thanks for the explanation Majenko!

Do you have experience with the wiring that you use to guide 5A current? I mean, I used these standard breadboard wires, which seemed not to withstand either the short circuit, or the overcurrent due to startup. Maybe I should use different wires?

Thanks again you all for your answers!

For those who are interested, in attachment you find the setup. Bare in mind that this was just a testing setup, so it does not look clean and slick, but then again, that might also be the cause of the issue in the first place :wink:

In the second picture I indicated the position of the melted wire. The USB cable was connected from the Arduino Mega to the PC (not on picture).

EDIT Paul (below) correctly noticed that I connected the + of the power supply to the - of the RAMPS board. Of course this is a recipe for disaster. This was not the case in the original wiring.