Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Uh Oh - Da Vinci crash / black bars
#1
Been printing a lot the last few days. I finished up a 20 hour print and almost immediately started an 8 hour print.

During the 20 hour print I noticed the extruder temp was creeping downwards a bit. I compensated by raising the set point. A set point of 228 yielded an average temp of 225. Normally, I set it for 230 and get 230, 225=225, etc. So I thought perhaps the heater was starting to take a crap again. During the 8 hour print my suspicion was confirmed. About 6 hours into the print I lost all heat on the extruder. I tested and had voltage, and the heater cartridge was an open circuit.

So, I Allowed everything to cool down. In order to remove the old filament once the new heater was installed I powered the machine on and used octoprint to set the temp to 200. In reality, the temp wouldnt matter because only the heater was plugged in. Nothing new- I have done the same thing several time and just manually plugged/unplugged the heater. Anyhow, While power was applied I unplugged the heater cartridge. brushed it off, and plugged it back in.

Im not sure what happened, but when reconnecting the heater cartridge (again, power was applied to heat the extruder), the printer flipped out - The LED lights went out, the LCD went to the two black bars, etc. It seems very consistent with a firmware erase from JP1.

So I thought.

I am unable to get any windows computer (tried winxp, win7, and win 8 64 bit) to recognize and accept the driver. I'm not new to the drivers, obvious doing development, so this is very puzzling. (I was originally -without luck- using a computer that was previously used to flash the chip via bossa.)

It seems some form of damage has occured to the atmel chip. I have checked the board and the power supply is intact, everything says it should work. I get the windows chime when the board is powered up, but absolutely can not communicate with the chip via usb. It appears that some sort of glitch has cooked the atmel chip.

Yea, I know, dont hot swap components. But really, That shouldnt have effected anything considering I had a heater failure with an intermittent connection that was causing it to constantly connect/disconnect every time the head moved.

Bummer. Might be a few weeks before I can get a replacement chip and try swapping it.

Im open to other ideas before I swap the chip.

Anyone else had anything similar??


Kieth
Reply
#2
looks like this issue no ?https://forum.voltivo.com/showthread.php?tid=8250
Reply
#3
At first glance it seems similar - but I am not able to get windows to recognize the board at all. His board is showing up as a bossa. On mine It shows up as an unknown device and no driver -bossa, etc- allows it to be seen or recognized. Ive already uninstalled all the root hubs, cables, etc with no change. I can program an uno board without problem.

And his description of the lights flashing and then it happening - thats exactly what mine did. Maybe the connecting of the heater cartridge was coincidental.

Ive got a mega and ramps here, maybe I will consider a swap as dont really use the on board LCD/ switches for anything.

Kieth
Reply
#4
I have finished examining everything possible and spent hours practicing google-fu on this thing.

It appears it is possible there can be a power glitch which calls the bootloader, potentially corrupting the chip. It is dependant on Brown Out Protection, and Im guessing the board BOP isnt the best.

I have a cable on the way to attempt to program via the UART which is an unused connector with 4 pins nears the bottom right of the motherboard. some parts may not be populated depending on board revision.


I read mixed statements regarding the bootloader on the sam3x8e chip. Some say it is programmable, some say it is ROM programmed by atmel and can not be changed. Anyone know for sure?

Kieth
Reply
#5
Summary: Something killed the chip. It is definitely dead.

Weighing some options -
1) I can replace the chip. The issue is without understanding why it failed it may fail again, or I might find something on the board is bad and killed the chip--- thereby wasting time and money with the chip
2) Get lucky on ebay and score a working motherboard. This seems unlikely, since I have only ever seen one board listed.
3) Install a mega/ramps setup.

Whichever route, Im not giving up. Frankly, Id buy the printer again, I dont think there is anything comparable for the price point...

Kieth
Reply
#6
I feel your pain. I have not been able to find a board anywhere. I will be watching this post closely. Good luck.
Reply
#7
I've ordered the chip to swap. Should be here in a few days.

What's wrong with your board? I've got a few chips coming, so might be able to repair yoursr depending on where you are and what's wrong. (Let's see how mine goes first. It's not the challenge of replacing the chip, that's easy. It's the risk that something else killed the chip...)

Kieth
Reply
#8
I am the one in the user to user forum that has the same issue with the black bars and no connectivity.
Where did you order your chips? I have emailed XYZ at least 10 times since getting my printer and have gotten zero response from them.
Reply
#9
Where are you getting parts for the Davinci????
Reply
#10
Promise, I'll let everyone know all the details once i get the chips and verify i ordered the correct one. Smile

I'll also post some pictures, or maybe even a video, of how to replace the chip.

Kieth
Reply
#11
I'm not yet at the printer to verify 100% but i think this is it.

digikey.com

Kieth



[Image: 20141206_121558.jpg]
Reply
#12


[Image: 2014-12-1020.26.07.jpg][Image: 2014-12-1015.51.49.jpg]
Reply
#13




[Image: 2014-12-1020.26.30.jpg][Image: 2014-12-1020.26.33.jpg][Image: 2014-12-1020.26.45.jpg]
Reply
#14
The last picture is how the chip looks after the flux was removed. It went as expected, no trace lifting.

For those curious of how it is replaced: I removed the old chip with an SMD hot air reflow station. I place the new chip by hand under magnification, then tack in place with a regular soldering iron. Re-position as needed, then I solder with a technique where essentially a solder bridge is applied to all pins and is then removed with desoldering braid. The braid has a lot of flux (which is why it works). In using the braid, I keep the portion of the braid which is almost saturated on the pins I am removing the excess solder from - this allows sufficient solder to remain in place to secure the chip and connections. Then I cleaned the board with rubbing alcohol. You can kind of make out the various steps - some of the pics look like there are solder bridges, but that is just the flux on the board reflecting the light...

BTW, I used to change these without magnification. The joy of getting older. Smile

And by easy to replace the chip - I mean easy if you have the right stuff. No way are you replacing this without an SMD reflow station. Wink

So, now that the chips is done, does it work?

Sort of. I now have connectivity and have flashed Lucs .92

1)The up and down keys on my display do not work. I temp modded the firmware for the right key to be the down key to allow me a painful experience navigating the menus. This might not be a big deal, since I essentially never touch those keys. In any case, I need to review my work to make sure if it is a firmware, ribbo cable, encoder, or motherboard issue.

2) Lucs .91 constantly resets every (aprox) 10 seconds or so.

3) The thermistor reading are not in the ballpark - the bed reports it is 138C when cold.

4) I found the heater cartridge cause the 0 ohm resistor (used as a fuse) on the motherboard to let the genie out. The heater is shorted internally. So, it definitely appears the heater crapped and cause a power issue, which in turn glitched the atmel destroying it.



No printing yet, but I am not giving up on the motherboard just yet. Its getting late, I probably wont have time to do anything else until Friday or Saturday.

Kieth[Image: 2014-12-1023.04.54.jpg]
Reply
#15
Hi Keith
about : 2) Lucs .91 constantly resets every (aprox) 10 seconds or so.
it means you used variant.cpp which enable watchdog (possibly the 0.92 one ?)- it is not supported by 0.91 so far

wow very clean work :woohoo:
Reply
#16
Thank you.

I didn't think about having copied the .cpp file - yes it was from .92 so im sure that where the issue is.

BTW, thanks for your work on this, you have done many of the things i wished to do but just haven't had the time. I'm sure it will be much more safe with the watchdog. I'm also curious, but hope not to test, if the watchdog will eliminate the glitching of the mcu. Another btw, there are a handful of reports online of this atmel mcu glitching and wiping out the bootloader, rendering the chip trash. I'm fearful I'm not the first and won't be the last.

Kieth
Reply
#17
Yes very nice work. Please keep us updated. Im thinking my failure was caused by the heater board as well. Sometimes a slight wiggle of the board would cause the whole machine to turn off.
Reply
#18
Just realized my board has a removable CPU and yours looks like it is directly on the board?
Reply
#19
Yes, mine is reflow soldered directly to the board. I probably should have been more patient and hunted a socket version if possible.


Can you post pictures of your board?

Kieth
Reply
#20
Sorry just woke up and was looking at a board for my other printer that is dead lol. My Davinci does not have a removable chip. I hope you get your board to work and then you can fix mine Smile
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)