Also, if you have your heart set on proving me wrong, then use this
fairly safe test on the board. You need a FTDI cable, but really any
USB breakout will work. What we want is a current limited 5 volt
source of no more than about 500mA (AKA 1/2 Amp). You then completely
pull the board out of the machine and remove all wires. Power the
board via the endstop connectors from said 5 volt source. This is
insanely easy provided you have at least a FTDI cable or whatever
power source and jumper wires that are preferably female on one end.
The endstops ports (all of them) have 4 pins. The center 2 are ground.
The bottom pin closest to the edge of the board where the heaters
connect is the 5 volt rail. From the FTDI cable, you ONLY use the
black and red wires. Black is ground and Red is 5 volts. I then used 2
male to female jumpers from Adafruit oe Sparkfun to then feed the
board via the male endstop pins. Again the basic idea is to power the
board from a current limited source. Should the board be shorted so
bad, it will trip the overcurrent on that USB port. Once you have the
board powered many will still do the turnon beeps as the atmega1280 is
a damn tough chip to completely kill. But that is the problem, even if
it partially survives, every single board to date tested had the
serial port killed when the 8u2 fails. This more or less ruins any
usefulness of the machine as then the only way to flash the chip is
via the ICSP and a programmer, you cannot connect via rep-g to set
steps per mm or other settings in the eeprom. Yes, we were smart
enough to remove the 8u2 or the 2 resistors in place between the 2
chips for serial coms and the mega still is always hurt.
The obvious sign here are that even with the board hack powered,
plugging in USB will show the 8u2 is dead as the computer will not
recognize the usb. Further, begin doing the heat test while powered.
By that, I mean put your finger on each and every chip and see what
gets warm or hot. Trust me, you'll know the difference. Most all of
the other parts on the board are always damaged because the cascading
failure always puts 24 volts down the 5 volt bus. For whatever reason,
this should blow the mega up completely but as I said, I have seen it
partially survive but it's so damaged and unusable, basically trash.
On the one board we (and really it was James Armstrong doing most of
it) attempted to fix every single chip was replaced because they all
were shorting the 5 volt rail or drawing excessive current and
obviously not working AKA blown.
The following list:
5 digipots
Mega 1280
8U2
Logic chips for LED driver circuit
2 thermocouple amps
FETS for logic circuit
Logic for thermal cutoff
all 5 botsteps
Again, let's just add that up from Mouser or Digikey
5 digipots
Mega 1280 $18
8U2 $20 (because the replacement part must be reflow soldered a
Sparkfun breakout is easier but more expensive)
Logic chips for LED driver circuit
2 thermocouple amps $30
FETS for LED circuit
Logic for thermal cutoff
all 5 botsteps (minimally $55 for Pololus or a lot more work to fix
the original board and again this is a reflow solder part)
Sor for basic repair/testing, you are going to need at least the mega
and an 8u2 breakout (about $40). That's so you can find out what else
is blown, likely the $30 in thermocuple amps. So now that's $70 plus
shipping. We haven't even touched a replacement regulator and likely,
all 5 botsteps are blown so we still aren't at a working board.
Just so I'm providing more info than just complaining, here is the BOM
of parts that likely failed:
Makerbot MightyBoard REVE BOM,,,
1,LM1084-5V,LM1084-5V Regulator,U13
1,V_REG_MIC2920,3.3V Regulator,U14
2,AQV25,Solid State Relays,"U6, U7"
1,ATMEGA8U2-MU,ATMEGA8U2-MU,U3
1,ATMEGA1280-16AU,ATMEGA1280-16AU,U4
5,BOTSTEP17,BOTSTEP17,"BOTSTEP-A, BOTSTEP-B, BOTSTEP-X, BOTSTEP-Y,
BOTSTEP-Z"
2,MAX6675,MAX6675 Thermocouple Chip,"U1, U2"
5,MCP4018,MCP4018 Digital POT,"U8, U9, U10, U11, U12"
1,NC7WZ02,Dual CMOS NOR Gate (Used as SR Latch),U5
1,PCA9533,RGB LED Driver,U15
2,PGB1010604,VARISTORCN0603,"VR1, VR2"
3,FDV301N,FDV301N NFET,"U16, U17, U18"
1,"XC7WH14DC,125",Tri Schmitt Logic Inverter,U19
6,PSMN7R0-30YL,PSMN7R0-30YL FETs,"Q1, Q2, Q3, Q4, Q5, Q6"
On May 3, 11:41 pm, Jetguy <
barrych...@hotmail.com> wrote:
> John, while we certainly hope that would be the case, I can garantee,
> if the regulator failed, then the extremely voltage sensitive 8u2
> failed and thus, you have NO way to communicate with the processor via
> USB.
> Understand, this isn't a one off, many, many of these boards have
> failed and it is always a complete meltdown. When 24 volts goes down
> the 5 volt bus, every single chip is destroyed.
>
> 44 message long thread on failureshttp://
groups.google.com/group/makerbot/browse_thread/thread/6e81b02d...
> 11 messageshttp://
groups.google.com/group/makerbot/browse_thread/thread/26bc2aab...
> 13 more herehttp://
groups.google.com/group/makerbot/browse_thread/thread/c1f49273...http://groups.google.com/group/makerbot/browse_thread/thread/26bc2aab...http://groups.google.com/group/makerbot/browse_thread/thread/c1f49273...
>
> Then we go into failure analysis from failed boards:
> 66 messages just on that threadhttp://
groups.google.com/group/makerbot/browse_thread/thread/e67e07c1...
> At least one protection device herehttp://
groups.google.com/group/makerbot/browse_thread/thread/4184079f...