I woke up this morning with my 20 hour long print failed around 15 hour in. I checked the octoprint console, and saw a very weird temperature graph. Later found out thermal runaway protection kicked in, and I thanked the gods for having decided to install a custom firmware to enable that protection.

I have an anycubic i3 mega, recently upgraded with an E3D v6. To make that upgrade, I had to splice new connectors on both the thermistor and the cartridge, as the ones they came with weren't going to work.

Ever since upgrading I started experiencing some weird Z-wobble, but I am coming to the conclusion that it may actually be a slight consistent underextrusion. The final layers of the parts I had been printing, are all severely underextruded, going as far as delaminating. I have attached both a picture of the temperature reading, and one of the failed parts. You should also see the underextrusion I was talking about in the first layers, and the major underextrusion in the final layers.

Now, I don't have a lot of experience with this kind of issue but I think the temperature was jumping around too quickly to actually be a good reading. Sadly I was only able to take the final part of the graph before the printer shut down.

I have taken some steps to try to avoid this from happening again

  • I have reseated the thermistor and the cartridge
  • I have reseated the connectors for the thermistor and the cartridge
  • I have re-zip tied and changed the orientation of the wires and hotend to something that seemed better.

Before this all went down, I had ordered a titan extruder to try to mitigate the underextrusion, as I read online (I think it was from E3D official sources) that the internal pressure of the v6 may be higher than the standard hotend and therefore I may have needed a geared extruder.

Do you have any advice for me? And if you could also help me figure out if the first layers actually show underextrusion or Z-wobble, thanks for making it this far.

I should mention that this had happened to me already but without triggering the protection, the part did heavily underextrude but it was just a test part while trying to find out what was going on with the "Z-wobble." It doesn't seem to consistently happen every time I start a print, but I figure that long prints would certainly make it more likely to happen.

Temperature Graph Failed print

  • $\begingroup$ how tight did you make the thermosensor clamp? $\endgroup$
    – Trish
    Dec 12, 2019 at 22:39
  • $\begingroup$ Just as recommended, tightened until grub screw touched, and then an additional 1/8 of a turn. Also did tug tests. e3d-online.dozuki.com/Guide/V6+Assembly/6?lang=en $\endgroup$ Dec 12, 2019 at 23:43
  • $\begingroup$ When you installed the new firmware, did you PID tune your hotend? Also, is the part cooling fan blowing on the nozzle block? $\endgroup$
    – 0scar
    Dec 13, 2019 at 7:41
  • $\begingroup$ @0scar No, I did that after installing the hotend. The firmware was installed much earlier. The part cooling fan doesn't seem to be blowing on the nozzle, as I have done some prints at 100% fan speed without issue. $\endgroup$ Dec 13, 2019 at 9:57
  • $\begingroup$ If you had 100 % fan speed always on, then the hotend can keep the temperature, it is a problem if it schedules on/off constantly. $\endgroup$
    – 0scar
    Dec 13, 2019 at 12:46

1 Answer 1


Periodic temperature irregularities, such as cycling between a higher temperature and lower temperature slowly enough that it spends at least a layer or two at a different temperature than other layers has a tendency to be mistaken for z wobble. You can actually intentionally modulate print temperature (at least of PLA) every few layers to great a sort of banding or 'wood grain' effect. Higher temperature, for whatever reason, results in thicker perimeters and cooler temperatures thinner ones. This is when always staying above the normal PLA print temperature though. The temperatures in the graph would certainly cause serious under extrusion during the times it was under temperature.

Understand though, it is likely that temperature has been behind the underextrusion, rather than it being a problem in of itself.

It isn't possible for the temperature to jump around too quickly to get a good reading. The temperature changes relatively slowly, simply due to the thermal mass of the heat block. It takes a little bit for the extruder to come up to temperature, right? When warming up from room temperature, watch how fast the temperature changes. That's the fastest the temperature can change, because that is when the heater cartridge is on at full blast. And even then, it isn't particularly quick.

It also does not appear to be a problem with the thermistor. If there was a poor or failing connection (like a wire that was almost broken), this would cause added resistance to the thermistor reading, and thermistors usually lose resistance as they heat up. So extra resistance throwing off the reading would make the printer believe that the hotend was cooler than it actually was, and you would wind up printing at a too high of a temperature. This would cause a number of problems, but your interlayer adhesion would be excellent. Delimitation would not be one of the issues, nor would underextrusion.

Thermal runaway protection is kind of a misnomer, because it is really 'something about the temperature of something isn't behaving like it should' protection. Basically all it is is a hard coded temperature and time window. If it has turned on the heater, there is a hard coded number of degrees that the thermistor must increase by within a certain time window, usually 30 seconds to a couple of minutes. If it doesn't, something is wrong. Or, if it drops out of the intended temperature by another hard coded number of degrees for a certain amount of time, then again the thermal runaway protection will be triggered.

In cases where the connection to the thermistor is failing, this will prevent the hotend from being heated without any limit (poor thermistor connection means the temperature the printer reads is always much much lower than what it really is, so the printer keeps trying to heat the hotend up hotter).

But thermal runaway protection will also get triggered for much more benign problems, like a failing connection to the heater cartridge. If the thermistor is working fine, but the cartridge is not working correctly and either only heats intermittently, or has a poor connection and can't get enough current through it, this will also result in the temperature not going up as expected, thus triggering the thermal runaway protection.

Based on your symptoms, the thermistor reading looks to be quite accurate. What you describe is exactly what I'd expect would happen for a print where the temperature of the hotend really did vary exactly as shown in your graph.

What can often happen is one of the leads (especially close to the cartridge but really, it could happen anywhere along the length of the two leads) will have broken from repeated wire strain (if you have disassembled your hot end at all or otherwise disturbed the heater wires in anyway, this is more than enough to cause a lot of metal fatigue), but the insulation around the broken wire will hold it together such that wire will still be making contact with itself. But it will be a more resistive connection, and will cycle between an acceptable and poor connection as the print head position moves.

Long story short, you get a temperature graph that looks just like that, because the heater cartridge is periodically either developing a poor connection, or losing its connection entirely, only to regain it again as things move just the right way again.

I would double check all your connections to the cartridge, checking the actual resistance and not just using a continuity tester. If those seem ok, then you will probably just need to buy a replacement cartridge. I like to keep a few of them on hand since they have a tendency for their wires to break just from a little bit of normal manhandling.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .