Somehow killed 2 arduinos in a row

Hi all - my Prusa was working perfectly with an arduino mega 2560 and ramps 1.4 for weeks, then one day when I turned it on there was a sudden surge of current from the power supply, and the sound of the arduino disconnecting from my pc. Assuming there was some kind of short I immediately powered down and checked all the connections. Everything seemed fine, so I tried powering up again. This time the resistor adjacent to the USB port on the ardiunio spontaneously combusted. This was starting to seem quit serious. I took everything and started to check for shorts in detail. Not being able to narrow it down to any one component, I decided I needed to clean house and replace everything ( ramps, drivers, arduino and even the psu).

When all the new stuff arrived, I assembled everything but didn't yet hook it up to any of the motors/heaters/sensors. I crossed my fingers and powered on, only to hear the arduino immediately disconnect from the pc. No arduino fire this time, but I can't connect to it any more so I believe it is dead.

What an earth is going on? The only constant between the two tests was my pc! And why would a current high enough to melt a resistor be flowing through the arduino? could there be some crazy potential difference between the ground on my pcs USB and the reprap power supply ground? I can't find one with my meter... Also before anybody suggests it, no I didn't hook the power up with the wrong polarity at any point.

Any ideas would be most appreciated.

Thanks, Ben

Edited 1 time(s). Last edit at 06/11/2012 07:10PM by beentheunseen.

... you can have some weird potential issues with switching power supplies and GND-connections over USB - in one of my last projects I assembled ten identical testing/installing rigs with RS232<>USB on the same PSU and PC ... and two of the rigs occasionally blows the fuses ( 0.7 Amps) or disconnect/clear the LCD-displays of the attached units

There's no measurable difference in the rigs, but this two (and only this two) behaves sometimes as if they'll receive a massive surge ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Ok, think I found the problem! I was indeed getting a potential difference between my arduino ground and the PSU ground, because the PSU was outputting 0V on the +12v rail, and -12v on the 0v rail! Measured relative to itself it was producing a perfect 12v, but as soon as I connected it to my pc, boom, 12v went straight through the arduino to ground.

Has anyone seen this happen to a power supply before? It was a cheap one from eBay, but I couldnt have imagined it could fail in such a damaging way...

Ben

Edited 1 time(s). Last edit at 06/15/2012 08:01AM by beentheunseen.

I think I have accomplished this same thing. I believe I also killed two arduinos in this way. What/How did you measure the power supply/usb connectons on pc to tell what is happening, Im pretty familiar with voltages and testing them, but not exactly sure what happened to cause this by your description above? My setup did also work for weeks but all of a sudden failed. I thought it was the arduino and bought another one. After hooking the new one up to my pc, showed up fine. but then after plugging in power to my RAMPS 1.4 board, sound of usb disconnecting and arduino DOA. Can you explain any further to help me understand? I also bought a 30A 12VDC power supply off of Ebay.

Plug your suspect power supply in and turn the power on. Then plug a usb cable into your pc with your pc switched on. Set your multimeter to volts, and measure between one of the ground (black) connections on the power supply and the ground pin in the usb cable (look up the ground pin here: [pinouts.ru]). The measured voltage should be close to zero. If it's -12v then you're looking at the same problem as I had.

I was then able to measure the voltage between +12v on the suspect power supply and 0v on my pc, and found this to be zero. This meant that relative to a healthy power supply, the dodgy one was outputting 0v on the 12v rail and -12v on the 0v rail. No problem if it's opperated by itself, but as soon as it's hooked up to another healthy ground, theres a 12v potential accross very low resistance and thus an epic current surge through the ardiuno.

Which power supply did you buy? Think mine was one of these: [www.ebay.co.uk]

I replaced it with a more expensive one and now it works fine. I guess you get what you pay for...

Edited 1 time(s). Last edit at 07/13/2012 10:06AM by beentheunseen.

Many years ago I bought a cheap Yum-Cha special (as they were known) PC in a tower case - I think it was a 486, which shows how long ago it was. I had constant 50 Hz hum problems from the audio. Eventually tracked it down to the power supply. Inside the shielded box with its "no user servicable parts" sticker, the power supply PCB was screwed to the metal case on standoffs. One of the standoffs had trapped the mains neutral line against the case and torn through the insulation.

So the whole power supply case was at neutral potential instead of ground. Never did work out why it didn't blow all the fuses, but I guess the neutral potential was close enough to ground not to generate much current. I lived on the 3rd floor of an apartment block, so I guess the mains earth wasn't exactly low impedance. I fixed the short and all my hum problems went away.

So maybe the cheap PSU has something similar wrong inside it's "unservicable" case? Like, the 12v positive and negative connectors wired up wrong?

wow, wad an episode, to me ive been fiddling in electronics for many years, anything that goes above 12V 1A (12watts) in small packages can burn stuff pretty quick ... the 12watts have to go somewhere

when i was looking at the PCB heater beds, i was quite prohibitive about the way it should be run ... too high a wattage me thinks in many ways.

Well, if you show us how to reach sufficient temperature - about 120 °C - with less wattage, we're all ears

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Taking a snip out of one of your comments i.e. "Ok, think I found the problem! I was indeed getting a potential difference between my arduino ground and the PSU ground, because the PSU was outputting 0V on the +12v rail, and -12v on the 0v rail! Measured relative to itself it was producing a perfect 12v, but as soon as I connected it to my pc, boom, 12v went straight through the arduino to ground.

It did't go to ground Ov it went to -12v

You basical stuffed 24v up the Arduino on 5v line usb would have recieved 17v respective to Arduino ground. The Ov from the output of PSU is floating and all grounds therefore need to be tied together.

Only posted for understanding