Why would my hotend shut off after 5 hours into an 8 hour print?

Hi,

I let an 8 hour print go overnight. Watched it for an hour and everything was fine before I went to bed.

Got up 5 hours later to find the printer doing it's motions, but nothing coming out of the hot end. The extruder was trying to push filament, but obviously it was just clicking.

I noticed the hotend was turned off. I was able to turn it back on in Pronterface, but at that point I shut everything off and went to bed since the print was ruined anyway.

So what are some reasons that a hotend would get shut off late into a print?

I'm running RAMPS1.4 with 3 high flow fans on it and a 0.35mm J-Head hotend printing 3mm ABS. Power supply is a 1200W medical grade so that is definitely not the issue. The electronics should be plenty cool since the heatbed was still on. What gives?

Thanks,
Dave

Edited 2 time(s). Last edit at 12/20/2012 10:00AM by iLikeWaffles.

This is really frustrating. I just switched to PLA today and was doing a 9 hour large print no problem with no lifting finally.

Then 7 hours into the print the hotend shut off. I remember back in the day my heatbed would shut off after a minute, but it was because the fuse or transistor was overheating on the RAMPS board, which I remedied with fans.

Since I switched to PLA I'm printing at a much lower temp.

Any ideas?

what kind of firmware are u using, i know from marlin there is a lower temp and max temp indicator, but it w(sh)ould shut down the entire print. But maybe there is a flaw in it. but maybe just check it. If i remember correctly it was set pretty high as it was set for abs use lower th min temp to 150 or so and check again. what temps are you printing with?

else tr another hostsoftware like repitier, wich imho is lot better then pronterface. \i switched 6 months ago when the new release came out and im never going back.

Thanks for the suggestion; I just downloaded it and will give it a shot.

I was using pronterface with slic3r 0.7.1, which is an old version.

There are only two things that can turn my hotend off: the electronics or software right?

I will try a long print with repetier and see what happens.

Thanks,
Dave

yeah i was going to say check your wires tuning to your temp thing if there is the slightest bit of current getting thought it fucks up the whole machine but it should stop everything not just the hot end

So I tried printing the same 9 hour object again in pronterface and again the hotend shut off at the same exact time. At the 6.5 hour mark to be exact. I was watching like a hawk so I quickly turned it back on, but there was no message in the pronterface gcode log that said it shut off. The print finished fine after I turned the hotend back on.

Weird. Especially since it shut off at the 6.5 hour mark both times.

I tried repetier, but after printing the skirt everything goes haywire instantly. After the skirt the coordinates get messed up and it tries printing off the print area. Appears as though the y-axis gets offset by a lot right after the skirt. Tried it a few times too with same result. Different issue than my original.

Edited 1 time(s). Last edit at 12/23/2012 08:51PM by iLikeWaffles.

Look through the gcode to see if it has any temp resetting codes around the 80% mark.

My current setup has been printing for days. My only problem is filament getting stuck.

I have the same problem, only mine does it sometimes at the beginning of a job and sometimes 3 hours into it. Only happens with jobs which contain large trays of multiple objects. Never had it happen with single objects even if they take many hours to print. Using Pronterface and Slic3r 0.9.8 on Mac.

Hello I am also experiencing this issue

The temperature cuts off at the same spot each time. I have checked the gcode and there are no commands to change/kill the temperature at the point where it gets cut off.
It is not a wiring issue as all you have to do is send the command to set the temperature again.
This has happened with both pronterface and repetier with 2 different versions of slicer.
Sprinter firmware.

Any fresh ideas?

Only thing I can think of is to set an audible alarm somehow to warn you when the temp starts to drop. Unless there is some way to set a script to constantly bump the temperature setting where you want to keep it.

You could add a post-processing script to slic3r to reset the temperature every Z move.... bit kludgy, mut it might help.

STF for "slic3r post processing" for examples.

-Rob A>

I use pronterface (win7). I've noticed similar problems, but usually very early into my print - like, when I first start it. The issue is somehow correlated to how long pronterface has been open. I say this as I have this problem if pronterface has been sitting idle for a couple days, and then I go to print something. The hotend starts to go up in temp, but then starts going down. I've had this happen 3 separate times. Connection are fine, power is great.

What solves it for me is a reboot, plain and simple. If I reboot the host pc, I can come back, restart pronterface, and everything works great. Now, I have not tried a 6.5 hour print - most of mine are less than 5 hours. So, I can't say how long a fresh reboot will give you - maybe 6.5 hours is the limit for some reason. I think I'll start watching threads/memory/etc for the pronterface process - to see if it they grow over time even while doing nothing.

Also, and I'm not sure if this is related, but I can say that during setup, when I'm jogging the axis around (I don't use endstops yet), the responsiveness of pronterface is much lower if it has been running a long time. As in, maybe 1 to 2 seconds after I click 'jog x +10" does the actual print head jog. Rebooting solves this too. I'm starting to think both are symptoms of some memoryleak or other, similar issue.

I have turned on temperature smoothing in sprinter. I have had better results with not a lot of testing.

The reading was "noisy" and possibly jumped up past the 275 max and triggered a failsafe. I have completed 3 prints since then.

If I have time tomorrow i will try again with a specific gcode that killed the temperature in the same spot multiple time.

I had my hotend turn off 20 mins into a print (Mendel90) and again at exactly the same time when I tried again. I re-sliced the model (same settings) and it worked fine after that. I assume it was some kind of rouge G-code or a bad combination of moves or something! Hasn't done it since anyway...

Guys (and gals) I very strongly recommend eliminating the temperature monitor in sprinter (use at own risk). I never have these problems since, and more importantly, it does bounce around forever trying to hit 1 degree C accuracy, which the system isnt even capable of as the internal and external error may be around 5 degC practically.

At the 6.5 hour mark, you may have a dense infill region, perhaps this is causing the heater to accumulate heat overtime and basically set off the fuses? What kind of amerage are you using? From what I've seen, it wouldn't be impossible for the self-resetting fuses to trip, and ramps does act funny at times.

Personally, I think it is software not hardware. Check the firmware and also, are you running pronterface?

Have you tuned your pid values? If your integral term is very small the integration term will build up gradually over a long period of time - then if your hotend is more sheltered it will take forever to count it's way back down. The derivative term can do funky things if it's excessive too! Your hotend should be insulated to help it maintain a constant temperature.

I tried turning on temperature smoothing in Sprinter as suggested, and the mystery dropping of extruder temperature stopped. I was noticing sudden jumps in the sensed temperature, which I figured triggered the whole thing to shut off. Now I have a slow sine wave pattern showing on the extruder, but it's still showing 10 degrees of difference either high or low. Anyone here know a way to make it not do this? You would think that the thing would be able to compensate faster than that!

OMG Integer error, that has to be it

More evidence for relative position.

Simba Wrote:
-------------------------------------------------------
> OMG Integer error, that has to be it
>
> More evidence for relative position.

Not sure what you mean.