dead arduino mega (I think)

[JKensai]

hi there - I just finished building my ToM with heated ABP. Awesome
machine, lots of fun to build... great job to MBI for putting this kit
together for the community. I ran a few prints that came out amazingly
well.... very happy with the results.

Now, the bad news....

The very next day I switched on the machine with nothing connected to
the usb port on the mega. I then spent about 20 min fiddling with
files that I ultimately wanted to print. Suddenly I heard a snap sound
(pretty clearly the electrical snap sound one usually never wants to
hear) and a smell of burnt electronics (that usually one never wants
to smell). Of course I raced over to the machine, turned the power off
at the switch and unplugged it from the wall. Even before opening it,
I could tell that both the arduino and the motherboard smelled fried
(this is an old pc skill of mine: diagnosing by smell, not burning out
boards!) -- and after I opened it it was pretty clear that the arduino
smelled worse. No visible scorch marks. I am meticulous about my
wiring so nothing shorted out based on bad or stray wiring.

I disconnected the motherboard and the arduino from the ToM to see if
I could connect it to ReplicatorG alone. No luck. But when I connected
the mega it seems like the power autosense on the arduino did it's
thing and powered it off of usb -- for a few seconds, then the pwr led
would fade out... This coincided with the serial port being recognized
on the pc and then dropped as the pwr light faded. I noticed the chip
(right next to the dc power port and two up from the usb port) got
pretty hot. I pulled the usb cable before I damaged my pc (just in
case). Wall power to the arduino did not do anything at all. So, it is
pretty obvious it is fried imho.

Finally, I reconnected the arduino+motherboard to the coolermaster
power supply... the power supply came on for a sec and then powered
off. I'm guessing this is because the arduino is not coming up. Ok, so
that was all background, here are my questions:

1. could I have had a defective arduino that just blew up like this?
I've not heard that happening before...
2. although it was on a protected line, could the coolermaster
*itself* have surged to blow itself and the arduino out?
3. is there a way to check the health of the powersupply and
motherboard? I'm pretty sure the arduino is shot but I don't want to
find out the hard way that those other two components are gone too
when they kill my new mega
4. last but not least, I'm suggesting a surge from the coolermaster
killed the arduino but is there another more probable cause? I
quadruple checked the wiring and all is a-ok... could powering the ToM
up and leaving it unplugged from the USB cause some weird problem?
could a short in one of the endtop switches kill the board or
something (they all seem fine to me and light up when they are
supposed to).

Anyway, thanks to all for your help. I'm bummed that my ToM is non
operational already but I will get this thing back on its feet and
printing soon

I'm grateful for any help you can lend.

thanks, Jim

[Jim Karkanias]

As a quick update for for the community, here’s a pic of the dead mega.  While looking for scorch marks, I missed the obvious damage shown in the attached. It seems pretty obvious now what made the snap sound!

 

I would like to pause here to highlight just how awesome MBI support has been.  Since my wiring is perfect :-), I figured that this just my bad luck and the PS puked ... so I quickly ordered replacement parts from MBI but support determined that the arduino was likely defective and could've taken out the mb on it's way to the grave.  So, without my asking for it, they gave me an RMA and offered to reimbirse me for the replacement parts!

 

I have rarely experienced this level of excellent customer support and while I was already a fan of MBI, they now have a supporter for life and I've already told my engineering friends about this (to which the most common response is: "hmmm, maybe I should take a look at getting one of these kits too").

 

I will be testing the PS (with a PS tester) and will reassemble the ToM as soon as the replacement parts arrive.  I hope to be able to report that I'm back online and functioning by the 18th or 19th or so and will report what I find.  My hope is that this whole thing is due to a bad mega as MBI and I have both diagnosed and/or the interaction with a failing PS.  Remember, I ran the ToM for a few moderate prints (10 min of of plastruder run-in, calibration cube, calibration tower, knurled bolt with matching nut, and a pretty complex buddha...) so about 2 hours of run time at least.  if bad wiring was the cause, e.g. some short in the endstop, or something weird on the safety cutoff (mk6+ extruder), I would have expected a failure right away.  Not 20 min after a power on the next day.  In any case, any thoughts are appreciated on this.

 

thanks, Jim

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To post to this group, send email to make...@googlegroups.com.
To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.

[Bill Culverhouse]

Obviously too much magic smoke in that little piece.

 

-b

[Jim Karkanias]

Indeed!  My experience so far has been limited to letting the magic smoke out through my own actions.  I know MBI employs cutting edge wizardry so components with an extra dose of magic smoke was an inevitable improvement

---

From: Bill Culverhouse
Sent: Thursday, June 09, 2011 12:55 PM
To: make...@googlegroups.com
Subject: Re: [MakerBot] dead arduino mega (I think)

[Jim Karkanias]

Okay, just a quick update here . I wired in a new MB + mega courtesy of MBI.  Did I mention again how great they are?!

 

I'm back up and running and everything seems A-OK!!  Prints are awesomly great; really happy with results.

 

I did have a small scary moment when I'd forgotten to invert one of the axes (the new motherboard obviously didn't have my previous settings), so the x platform headed in the opposite direction of the end-stop. I looked at it stupidly until it crashed into the case wall (naturally). the stepper strained for 2 sec while I tried to figure out if I should kill power our just hit stop on the build. I killed the power. Here's the scary part: the PS would not come back on 10sec later. I unplugged everything, made sure my surge protector didn't blow up, etc. Plugged everything back in about 2 min later. Still nothing. By now I was thinking the PS really was trashed and was feeling sorry for myself. :-)

 

Eventually, I tried the switch one more time for some reason and it worked. and has been working since... I'm going to keep an eye on the PS and test it with a digital PS tester when I can get my hands on one and have the time to crack open the bot again. I will let the community know what I find.

 

Both Ethan (MBI support) and I believe that the thermal fuse in the PSU is the likely culprit. All three steppers were going (the system was finding zero and each axis was moving towards the endstop -- except for X which was going the wrong way that is), the extruder was heated to 220 after having just run a 5 min extrusion test, and the ABP was heated to 120 so I think it's safe to say that the PSU was fully loaded as the only thing that wasn't going was the ABP drive motor.  I'll calc the load but does anyone know how much the bot draws full out under these conditions?

[David Rosen]

Regarding the power supply not turning back on: TOM seems to be very finicky about being shutdown properly.

I have the same exact problem that you described:

1. Crash one of the axis into a wall or stop switch

2. OMG! Hit the power switch

3. Bot will no longer turn back on. 

4. Wait 10 minutes, then it will come back.

I've had this happen 3 times. Also, improperly yanking the USB or putting my laptop to sleep while connect to the TOM seems to cause similar issues, but not 100% of the time.

Now I just hit the Stop or pause button in RepG and move the platform away from wall/limit switch, then disconnect RepG, then unplug the USB, then shutdown the power. No more problems.

[B Stott]

There are safety cut-offs within the Power Supply and on the A3977 controller chips. They may save your bu... er ... bot.

Other  more knowledgeable will have to pipe in for all the safeties of overload, thermal, ground, reverse current, sag???

[James McCracken]

Typically to reset this condition (on my cupcake, similar technology)
I do the following:

Disconnect the FTDI cable
Turn off the power switch on the mother board
Turn off the power switch ON THE BACK OF THE POWER SUPPLY
Turn on the motherboard - watch as the LED lights green
Wait for the LED to unlight
Wait another 10-15 seconds to drain the rest of the residual charge
Turn off the motherboard
Turn on the power switch on the back of the power supply
Turn on the motherboard

The above procedure works to clear a blinking debug light... my bot
has spent maybe 85% of its operational life with the debug light
blinking as without this procedure, you have to unplug it and wait
30-60 seconds (minimum) before plugging back in to clear the residual
charge and power down the motherboard completely.

If you didn't catch that last bit... let me be explicit. The only way
to turn off the processor on the motherboard and force it to reset and
cold boot is to remove power from the bot. Turning it off just
doesn't cut it.