Re: Do I need to leave my PC turn on when MakerBot is printing via USB?

[Jetguy]

Well, that depends.
If you wrote to SD card card, then inserted the SD card into the
Machine, and then used the print from SD card function, then you can
disconnect and shut the PC down.
If you just print from USB then no, the print will stop and not be
recoverable.

On Apr 30, 11:55 am, PolPPol Pol <polp...@gmail.com> wrote:
> Do I need to leave my PC turn on when MakerBot is printing via USB just
> like many other 3D printer?

[Jay]

Pol,

Per most of the 'heavy' guys on here...print from the SD card. USB can allow too many problems that can kill your print. If you have to use USB then you'll not only need to leave the computer on but you'll need to check your power settings to make sure it's not set to go to sleep after a certain time or sleep the hard drive. I had a print fail for reasons like that. There's some technical info from people like Jetguy (response above), dnewman, Eigthy, and Drone Pump that can give you specific reasons...I just know they said it's bad...so I don't do it anymore.

Jay

[MacGyver]

As one of the "heavy" users on here I've never actually got the print from SD card to work.  All 511 things I've posted on thingiverse + the 100's of other things I've printed have been solely printed via USB.  Do you guys actually take the SD card out of your printer and put it in your PC to write to it then sneaker foot it back to the printer?  Seems like too much trouble to me.

[Dan Newman]

Yes. Keep more than one SD card so as to cut down on round trips.

Dan

[Jay]

Yes, specifically bought a computer with built in SD reader...and it sits only 6" from the printer...so I don't 'sneaker' it much :)

and I've had errors occur when printing from USB that just don't happen using the SD....

Jay

[Bryon Miller]

If you are printing from the SD card AND your machine is plugged into the USB port as well, you can still shut down the computer and the printer will continue to print.  I have done this many times.  However, if you started the print from replicator g or something and printed directly from the computer without putting  the file on the SD Card, then the printer will stop receiving data from repG when you turn the computer off.

[Bryon Miller]

I write to the SD card like that.  Move the card from the machine to a card reader, copy the file and your done.  It's fast, takes seconds and you're done.  It took me forever to figure out why makerware couldn't see my rep 2 but replicatorg could.  It turns out that Makerware can't read your device if you plug it into a USB 3.0 port but RepG can.  I had to use the card at first because the software wouldn't work for me out of the box.

[tunell]

I ALWAYS print from SD. I've had too many troubles trying to print from USB, though admittedly that was with earlier generation makerbots.
What is unfathomable to me is why MBI didn't put on some onboard memory on the motherboard so that the full x3g file could be sent to the onboard buffer over USB and then run from there transparently, like a normal printer does.

[Bryon Miller]

I picked up a kodak card reader at Radio shack for $9.99 and it works fine.

On Thursday, May 2, 2013 11:23:53 AM UTC-7, Robo wrote:

When I first got my Rep2, I tried to print via USB and every print failed eventually with "serial error".  Having that happen 7 hours into a Yoda head isn't cool.  So I do everything on the SD card now.  Thankfully, the included one from MBI is quite large.

Another thing that I do, probably just out of paranoia, is to not leave the SD card fully clicked in unless I'm going to print.  The assumption is that I don't want to add undue fatigue to the springs that let it 'click' into place.  For about $50 I got a great little card reader/writer from Newegg that I use to copy x3g's to the card.

[Bottleworks]

That's an interesting point... On the rep1s, MBI had an unpopulated sd card spot on the Mightyboard.... I wonder if they were thinking of doing just that.

[Dan Newman]

> Another thing that I do, probably just out of paranoia, is to not leave the
> SD card fully clicked in unless I'm going to print. The assumption is that
> I don't want to add undue fatigue to the springs that let it 'click' into
> place. For about $50 I got a great little card reader/writer from Newegg
> that I use to copy x3g's to the card.

BTW, on a Rep 2, it's also a good idea to have an SD card in the slot. Owing
to it's orientation, it's easy for debris to fall in, including fine wisps of
PLA filament. I've helped fix three Rep 2's whose SD cards weren't working
owing to debris that got down into that slot.

Dan

[Dan Newman]

On 2 May 2013 , at 9:22 PM, Bottleworks wrote:

> That's an interesting point... On the rep1s, MBI had an unpopulated sd card spot on the Mightyboard.... I wonder if they were thinking of doing just that.

And I keep on meaning to mount a microSD card holder on to my Rep 1 board
and see if it works. It would avoid running over that ribbon cable….
From looking at the schematics, all the correct traces are in place.

Dan

[delsydsoftware]

I used to print exclusively using USB. I just recently switched to using SD cards. I had started noticing a lot of pauses during accelerated prints. My computer isn't a slouch by any means. It's a AMD 8365 clocked to 5ghz on water cooling. But, there were still issues getting commands down the USB cable. Also, some video games (Kerbal Space Program, I'm looking at you) would tend to screw up prints that were going in the background. Once you get used to the process of queuing up prints on one SD card while slicing models on another SD card, it becomes more efficient.

The reason I stuck with USB for so long was that the 5.5 firmware that my rep 1 shipped with seemed to have odd issues with SD card prints. I could print from USB with no issues, but the same model printed from the SD card would drive the print head into the platform. I haven't had that problem with sailfish, thankfully.

[Jetguy]

Overclocked to 5gHz. You do understand that overclocking, especially
bus speed changes are extremely bad for critical timing required for
USB communications?Then there is the RFI/EMI issue that overclocking
is sending huge amounts of radiated noise off of every single wire
both in and outside your computer. Then, couple that with the fact
people who overclock often have cool side windows and generally not
RFI/EMI tight cases with all the fan openings and lights and fan
grills. If that PC is in the same room as the printer, the ribbon
cable going to the LCD and SD card slot it being floooded with noise
from your overclocked PC. It's already known that the Replicator's
electronics design is sub optimal to the SD card slot but jamming the
crap out of it with EMI/RFI could easily explain your SD card errors.

Sorry, I'm not trying to bust on you for overclocking, just educate
you a tiny bit on what that does to everything else. I owned a PC
shop, I've built hundreds of overclocked machines. Only after I got my
engineering degree did I begin to see what overclocking does to
externally clock critical devices. Some folks are already claiming
they get temp errors when USB is connected due to EMI/RFI being picked
up on the thermocouples. I cannot imagine what your scenario is like
but very likely, worst case, for EMI/RFI.

Further, if you are running the latest Sailfish firmware, you should
use the SD card checking function which checks each read block of data
to ensure it is correct and uses the buffer to help with re-reads in
case of errors.

Again, I'm not doubting the PC is impressive, but I think you have no
idea how changing clocks affects other parts of the system. It's OK,
the overclocking forums should be more explicit in telling folks what
happens and what it affects. THe USB is affected because it is a
highly critical timing application. When you start changing bus speeds
(likely part of reaching that 5ghz) you actually slow down
communications with fixed rate systems like USB. Here's a shinging
example of a usb tuner being jammed most likely by bus speed and EMI/
RFI I listed above http://forums.anandtech.com/archive/index.php/t-2197464.html


If there was any tip I could give other MakerBot users, do not run a
modded computer anywhere near any MakerBot. CPU Speed, overclocking,
and mods do NOTHING for Replicator-G or Makerware to increase the
speed of the application but can and will affect operation of the bot.

[delsydsoftware]

Actually, I'm well aware of the timing issues associated with changes in bus and CPU speed. I've been overclocking PCs since the 486/33, so I have seen a lot of failures.  :)

 I did test the computer at stock speeds and had the same issue. That's generally my first troubleshooting step--go back to stock speeds and see if the problem goes away.  It may be due to the length of the USB cable I'm using---the printer is 10 feet away from the computer, and I'm using a 15ft USB cable to reach it. That's allegedly within the USB 2.0 spec, but probably is introducing some interference. I mentioned the CPU speed just to prove that it wasn't a case of a slow computer that couldn't keep up with the instruction queue.

[Eighty]

Well I, for one, am glad the over clocking discussion was had. I learned something today.

[delsydsoftware]

Oh, I forgot to mention that I'm actually using a stock 200mhz bus speed with a 25 multiplier. I could get more performance from a bus overclock, but I just got tired of messing with it. I played the bus-balancing dance with my last AMD setup and decided to go for easy-mode this time :)