Hi, I've just upgraded my OpenBeam Kossel Pro with an E3Dv6 hotend and am having some issues with 'Thermal runaway' reported. I am using Simplify3D software and printing from SD card. I have the 24v 25w heater cartridge. As per instructions, I have run M303, PID Autotune, and changed the firmware to suit and made sure the correct thermistor was selected. However, the problem would appear to be that the hotend cannot maintain the correct temperature and I get the 'Thermal runaway' reported after the software reports an under-temperature of some 5 degrees for a certain time. So although 'Thermal runaway' would suggest too high a temperature, in my case, it is too low a temperature and this is without any external fans running! It would seem to be that the autotune is not setting the PID values correctly. Any help would be really appreciated. Mike
The keyword is 25 w. Get a higher power cartridge, 40 w or so. The present one will try hard but doesn't get over the hump. Cheers, U.
I had wondered about the 25w cartridge but the autotune at 250C went fine and there was no problem reaching that temperature so why can't it hold at just 200C? I appreciate I am shoving cold filament into the hotend but the extrusion rate is pretty slow. What is strange is that the LED, that indicates whether the heater is on or not, is blinking quite slowly when I get the 'Thermal Runaway' yet I would have thought it should be on fully when trying to raise the temperature. That is why I was suspecting a wrongly set PID rather than the cartridge. Current PID values are P-46.65, I-5.16 and D-100.97
Also if you have the fan dircted at the hot end and it turns on full blast it can cool your nozzel too quickly. You want to slowly bring your fan speed up and or try to avoid hitting the nozzel.
Bad contacts, sensor correctly connected? Strange.Was then fan active during the PID-setting? Both the V6-own one and the parts cooler? The interaction of the parts cooler with the heat regulation is a mystery of its own....
As Ulrich says, the 25W heater is a little light in its ability to maintain the temperature but only just, whereas the 40W heater has enough grunt, provided your power supply can supply the current. The thermal mass of the hot end is quite small due to it being made of aluminium so fast print speeds can take more heat out of the block. Also, although thermal conductivity of aluminium is good, the thermistor is positioned such that heat from the cartridge may well be shunted by the nozzle and cause a greater drop at the thermistor. I have not had this problem but I have used a heatsink compound with both the cartridge and thermistor to ensure the best thermal conduction for both items. I also used the compound on the threads of the nozzle too and this makes nozzle changes easier i.e. no stuck thread. The term "thermal runaway" really covers both over an under temperature, which is a little strange, but we live with that. Mike
Thanks, Mike - it looks like I need to change to the 40w cartridge. I think the original, Kossel cartridge is 40w so, if it fits I will try it before ordering one from E3D.
OP: can you post your temperature graph, ideally with heater power shown also? What was the resistance of the heater when you tested it during assembly? 25 watts is plenty to maintain even 300C as long as you don't point a fan at the heater block. The heater power rating is not the problem here. While it is possible that a 40w heater may eliminate the thermal runaway, it is not the underlying cause. Also, mike01hu: thermal paste is not rated for the temperatures in the heater block. It is not recommended to use on the thermistor or nozzle, for this reason.
Unfortunately, I did not save the temperature graphs of any of the times I had the problem. I have since re-done the autotune and changed the PID settings again and, at the moment, the printer is printing okay. However, I have not tried to switch my part cooler fans on as I am pretty sure this would cause the runaway. The resistance of the heater was just about spot on 23 ohms. It is quite noticeable that on fairly long,fast travels at 150mm/sec, the temperature drops shortly after but regains whilst printing at 50mm/sec. It appears that the ambient air temperature is cool enough to have an effect when the hotend does rapid moves.
Stop action, until... - ... Mike F., until you have verified that your power rate settings in Marlin are not the pulled handbrake. If I am not totally mistaken and my memory plays tricks with me, there are some settings that define what percentage of the potential full power is applied to the heater (both nozzle and bed) during heating. That seems to be meant to protect the system against a local hotspot - where the cartridge gets very hot instantaneously but the surrounding material cannot transfer the heat quick enough into the provinces and last meadows. I am not sure if these parameters play a role in the PID parameter finding procedure. At the moment, I don't have access to Marlin sources here, but I am sure that you will find out or somebody else will know. Maybe a remainder from your previous setup? I had a Merlin hotend before the V6, and I remember that the V6 required some more umpphhhhh when I first set it up after the change. Sheeeesh, the world is sooo complex and non-linear...
That sounds familiar. I had to reduce that slightly to 160 or so for my 40 w V6-setup. Otherwise the initial overshoot - despite PID - was inevitable. For 25 w, it should be closer to the max value.
Elmoret, that makes a lot of sense - my current PID_Max is set at 125 and I was previously using a 40w cartridge. I will try changing this tomorrow and report back. Maybe this is something that ought to be in the instructions? UlrichKliegis, thanks for the heads up and prompting Elmoret's reply.
I am not so sure but it does need deeper investigation and the temperature graphs can give an indication. You are right Elmoret regarding the heatsink compound. I use a silicon compound without thermo-conductive fillers i.e. silicon grease, which is fine up to around 240C but I only print in PLA and its derivatives so the higher temperatures are not an issue. At higher temperatures it can convert into silicon carbide, which is an abrasive compound so there would be issues if put on threads that reach much higher temperatures. I should have put that caveat in my original post, for which I apologise but I was being rushed by the one who must be obeyed!!!. Mike
That means that you are actually driving the cartridge with about 12 w - not very much. I would be happy to hear that the problem could be solved this way. If not, we'll search on. Instructions - oh well, yes... hey, it's in the comments of Marlin!
Well, I changed the PID_Max setting from 125 to 200 and my hotend is now behaving itself very well indeed. It is now maintaining temperature within 1 degree either side of that set even with the external fans cooling the part. Thanks for all the help.
