[SOLVED] Repeatable Failure To Print -- Early in Raft

[EricPeterson]

The first time this happened I thought it was a printer hardware failure. While printing the support raft the printer just stopped in the middle of a line. It did this twice, continued, slanted a couple lines into one another, then finally ran over the edge of the print and drove into a limit.

I rebooted the PC, restarted the printer, re-exported the STL, ran the STL through a checker of my own design that I've used to verify my STL files for actual stereolithography for several years, restarted Simplify3D.... and repeated precisely the identical failure in the raft at the same location three times.

MakerGen M2 V3b, S3D 2.2.2.

The STL is exported from Solidworks 2014. I've been printing versions of this part for four days now... this is the third rev. It's the eighth or ninth little fixture I've been printing, after four test parts I designed and printed that were expressly designed to exercise printer performance. I've gone through a spool of PLA that came with the printer and half way through my first spool of ABS, and aside from clogged nozzles that prevent the first layer from going down properly, this is the first failure I've seen.

Here is what the failure looked like:

http://www.ericpeterson.info/M2Files/P1000238.JPG

Here's what the preview showed when it happened:

http://www.ericpeterson.info/M2Files/P1000239.JPG

Here's the STL and factory file:

http://www.ericpeterson.info/M2Files/10 ... tiwarp.zip

The M2 is happily printing right now the last version of the fixture prior to the one that causes the failure.

- ep

[KeyboardWarrior]

I've had this happen on my MakerGear M2 when printing a raft with a USB cable, but haven't been able to repeat it when I save the file onto SD card and then use the button in the Machine Control Panel (print from SD card).

[PPLDC3]

Glad I'm not crazy. I posted about the same thing recently, which has been happening on my TAZ 4 and my Airwolf HD2x, repeatably, and with both v2.2 and v3.0, and the error has occurred with a variety of different STL's. The only fix I've found is just unchecking the raft, but sometimes I really need it.

It definitely sounds like the issue is caused by a loss of connectivity. Sometimes, the line that shoots out to nowhere goes out at a different angle, basically continuing along the direction that it was going when the error struck.

[EricPeterson]

"A variety of STLs". Ugh. I was hoping it was a problem attributable to the regular array of island-like features in the first problem file, but now I've got the same issue with another one, and it's all sweeping curves in one big feature.

I'll try with the raft off. See how that goes. Yeah, the preview on the screen shows good paths... but what happens at the printer... very wrong. This time I wasn't watching it and I came back to a gob of PLA piling up in one place.

[PPLDC3]

I managed to fix my error. It has to do with the fact that while a perimeter is printing, the computer seems to periodically prompt the printer for a status update on how it's doing, and on these really long raft perimeters, the printer is so focused doing that, that it waits until it's done before reporting to the computer. Since I had my first layer speed set to super slow, the computer was triggering a default communication timeout and assuming the printer had gone dead, when it was just taking a while to report.

Fixed this problem by going into the MACHINE options, and on the last tab, changing the firmware communication timeout from 10 seconds to 30 seconds.

See Pierre-Luc's reponse here: http://community.robo3d.com/index.php?t ... sive.5036/

[EricPeterson]

This all makes sense. I really needed a solidly adhesing raft today to anchor a very complex support structure for a skeletal, spindly part, so after reading this last post and the links I disabled the super slow raft base layer and turned the other raft layers up to 100%. It appears that my PC's USB timeout is also on the order of ten or fifteen seconds. As long as every line of G code executes in less than that, the print runs fine.

Thank you all for the comments and info.

-ep