Is the problem my thermistor... again?

[PhoenixNZ]

Hi All,

Just started up my Rostock Max v2 again after about 2 months downtime. I've only printed 4 small things and the fan shrouds since assembling and I had no problems with those. Today though, I tried to print a simple set of 5mm cube steps in PLA. Pretty standard stuff like this http://www.thingiverse.com/thing:24238. Decided to print 3 of them at once to test accuracy of different areas of the bed and stringing problems.

Today, my print started off fine - the extruder temperature curve started off a bit janky while the perimeters were being printed but I figured it was just because the head was moving a bit of old PLA buildup inside the nozzle. It soon smoothed out into a normal, smooth, stable PID curve. So I started to browse the internet while keeping an occasional eye on the print. I'm using Repetier 1.0.6, with the default recommended PLA settings from the Rostock manual with a couple tweaks. Summarised below.

Extruder: 195°C
Bed: 60°C
Speed: 30mm/s print (200mm/s non-print)
Layer fan always on
Slow down print if layer <40s, slowest speed 8mm/s
Retraction 7.4mm, 0.2mm Z lift

After a while I noticed that there was a bit more stringing and smell coming from the extruder. Normal I figured since the print was getting into the 3rd set of steps and there may not be enough time for layers to cool.

Then I noticed there was a LOT more stringing and the print smelled very hot. Sure enough, checking the temperature graphs shows a major issue. Here's an imgur link here for the image if they don't show up properly:
[img]http://i.imgur.com/TIW3OC4.png[/img]
[img]http://i.imgur.com/dC7UPjU.jpg[/img]

Is this a thermistor problem? This doesn't seem to match the wiring issue I had when I first assembled the printer. The thermistor itself was fine but my wiring being dodgy meant that the extruder temperature occasionally went to 0 when the head moved in a certain way.
A temperature graph of what that looked like can be seen here: http://forum.seemecnc.com/viewtopic.php?f=37&t=6003

I also tried manually turning on the extruder and moving the print head around the bed at Z=50 but that did not seem to replicate the problem:

[img]http://i.imgur.com/5vrjj40.png[/img]

[PhoenixNZ]

And now I'm even more confused! I ran a second print of just one set of stepped cubes centered and got the temperature graphs below. Whatever is happening happened from about the 4-5th layer to the 20th(?), or in other words mid-way through the lowest step, then the jagged temperature reading magically fixed itself and printed normally through to the end.

Overall quality of the print is noticeably lower than the first attempt with 3 sets of the model. Though that could be to do with giving the layers more time to cool

http://imgur.com/a/efBpP

[img]http://i.imgur.com/VErOZ0A.png[/img]
[img]http://i.imgur.com/1GHZCzj.png[/img]
[img]http://i.imgur.com/9OGGYmm.png[/img]
[img]http://i.imgur.com/DwvDB5u.png[/img]

[BenTheRighteous]

I don't know how a failure-mode thermistor behaves, but to me this seems like a loose wiring problem. Did you follow the suggestion in the guide to wire up the hotend with an LP connector? If so, are you sure all the pins are secure?

I'd triple-check all wire joints at your hotend. This feels a lot like as your print head move a certain way at a certain layer, it bends your wire loom at just the right angle to loosen a connection.

You might also try turning on the heat while not printing and the head is stationary, and watching the temperature graph while putting a (small!) amount of stress on your wiring at various angles. Whether or not you can replicate the problem that way would tell you more about where to look.

[teoman]

I only print from the sd card. But judging from the green graph below. It seems not to be sending power to the hotend.

[Polygonhell]

Looks like either bad wiring or worse a loose thermistor to me, one of the worst failures is when the the thermistor creeps out of the hole, it gradually reads lower temperatures as a result the amount of power supplied to the hot end is increased, usually resulting in a destroyed PEEK section. It's how I lost my first hotend and there is almost no way for the firmware to protect against it.

[BenTheRighteous]

I just took another look at the first image you posted, and the very tail end of the graph looks very bad:

[img]http://i.imgur.com/TIW3OC4.png[/img]

(Note that I had to right-click and "view image" to see the whole graph.)

Looks like when your thermistor kicked back in, it was reading 270 and if we could "fill in the gap" before that, it probably hit 280 or more at some point.

I don't know at what point you start to cause serious damage to the hotend, but the firmware will prevent you from setting anything higher than 247.

[PhoenixNZ]

Thanks for the feedback guys.

Yeah; the images show up on a larger screen but probably aren't the best for mobile users etc. Encountered that problem before which is why I try to incluce imgur links if I remember.

That hot end has hit near 300°C a couple times. Once when I had the original problem, and again a few days ago when I posted this.

Guess it's time to pull it apart and re-wire.... again

[BenTheRighteous]

Yikes. Thermistors have an upper limit; beyond that, they're permanently damaged. I don't know about the one SeeMeCNC ships, but the one that E3D ships is rated for 295C.