Thermal runaway protection does not 100% protect your printer from over heating

Thermal runaway protection does not 100% protect your printer from over heating.

We all know the issue with China-Printers. They often, even in 2020, come with firmware from 2014 or earlier. Security features like thermal runaway often not existing or disabled.

What does “thermal runaway protection” do?

In easy words: It monitors the temperature over time and makes conclusions. If the heater is on for several seconds, but the temperature stays constant, something is wrong. The printer shuts heating down and goes into halt state. Another situation is when the temperature is not rising after a certain amount of time. This mostly happens for the heat bed of cheap printers. When setting the non insulated bed to reach 100°C the last 20°C take, if at all, forever. At some point it takes minutes to raise the temperature by just 1°C and the printer pulls the plug and goes to halt state again. This also happens with removable beds as some point. I for example use a glass bed. When a print needs some convincing to be removed, I keep the printer at temperature and remove the glass bed. I wave it around until it cools below release temperature and once the print is finally off, I put it back onto the printer. This sometimes causes the printer to pull the plug and go to halt, as the temperature rapidly gets eaten by the cold glass bed and drops easy 10°C at the aluminum plate, where the sensor lives. Another way this system protects is when cable get loose or break. In that case the temperature will instantly drop to somewhat zero and the printer knows something is up and shuts down. Also when a heater cable breaks or a heater cartridge burns out, the printer will notice that heating is not effective it will shut down.

This sounds like a very secure system. But it isn´t. The main issue is that there is just one sensor the system has to rely on. If the sensor is returning wrong values the main board is just believing and acting.

These sensors are just temperature based resistors which never lie and always work. Well, yes. You are right. The resistors never lie.

So how can a thermal runaway still happen?

Quite easy. The sensor gets loose. Instead of sitting deep inside the heater block, it may dangle at the beginning of the hole. The result is very bad. Instead of getting the heat, it gets just hot air, which is significantly cooler than the heater block. Even without the part cooler not the hot end cooling fan blow air around that opening, the heat can easy be around 50°C colder.

But the thermal runaway would catch that, right? 50°C difference is for sure causing an error.

Again, you are right. It will trigger a runaway, when it happens mid print.

But when it happens in between, e.g. when changing the nozzle cleaning the tool head or just over time it won´t.

If you turn on the printer everything looks normal. The printer shows the normal 23°C for bed and tool head as usually. If you start printing the tool head will start heating. The air around the sensor will heat and everything looks normal. But there is one problem. When the tool head reaches for example 205°C for PLA or 220°C for e,g, printing ABS in reality the nozzle and the therefore the filament inside the tool head gets 50°C higher than the printer reports. So Instead of the wanted print temperature you get 255°C for PLA and whooping 270°C for printing ABS. The result will be clogging and parts of your printer will melt, even when printed with ABS.

But it gets worse. Usually the printer overshoots by 5-7°C when heating. In this case and due to the air cushion around the thermal sensor, the temperature change in the nozzle easily shoots over by additional 10-20°C, as the air cushion builds a nice delay when transferring heat towards the sensor. So for a short time your printer heats up to over 280°C.

Why do I write this? Well, because it just happened to me and at first I did not get what was going on. I got a clog, which I basically never get, so I did the worst I could do. I increased the temperature and somehow it did not get better, but worse. The bowden snippet inside the heat break was molten, which I was blaming to the normal overshooting when heating to 220°C for ABS printing. So I inserted a new heat pipe and it burned out instantly. I got a electronics meter with a temperature sensor and it showed 220°C when the head temperature on the printers display was showing 170°C. It did not make any sense until I noticed the little thermal sensor bulb was sitting at the entrance of it usual position and not deep inside the heater block. Then suddenly everything made sense.

After repositioning the sensor the printers temperature and the external thermometer showed an identical temperature and the printer works fine again.

So for the future I will add a new point to my frequently printer inspection. Cross reference the temperatures with an external thermometer to see if everything is in proper operational state.

PS: I did not know a heater cartridge can actually glow. Turns out it can! Very bright! And it still works! :smiley:

Good post @Geit. A fair warning to everyone. I also think it’s a point in favor of the screw-in thermistors that @roon4660 likes so much.

I’ve been trying to get over another episode of brain failure. My Ender 3 pro is down for the moment because I can’t make my tiny brain concentrate …again.

For some unknown reason the chord to my printer shorted out. I am just not in the state of mind to do all the obvious things that I should do.

I go through this crap all the time but I will return. I need to trouble shoot the power source and find out what is happening. I am just waiting till my brain is again a little bit

functional.

Sorry to hear that. I hope you feel better very soon.

Very difficult to explain to anyone that has never been struck by lightning. I’ve had 21 years of this crap but in a certain sense it makes you strong.

Sad to hear. Hope you get better soon!

It’s more or less routine by now. I’ve been trying to get up the mental energy to deal with fixing this thing and installing the new fan.

Question about thermal runaway.
Does the Smith3D Firmware have all this logic in it to deal with thermal runaway?

I would be very surprised if it didn’t. It can always be enabled if it isn’t. The pros here can help you with that. I always have trouble trying to get it to quit giving me error messages that no one can explain to me.

Yes it does someone on their site asked the same question and the answer was yes.

It took additional 120 hours of printing and the heater cartridge finally died. Luckily just in the first few minutes of a print and not during the 8 hour print before :smiley:

This time thermal runaway was triggered of course. :smiley:

Well, that was a break. Do you normally keep a spare heater cartridge on hand?

I usually have a couple of everything to spare, as I usually buy more than one, when I need a part. However I never ordered heater cartridges, but I ordered a new E3D hotend clone when the first runaway happened, as the block was bend, the fan was crappy and everything kind of rocked down. So I swapped that one in. I still have the original cartridge, which came with the ANetA8, which should be next to new with only 20 hours of printing or so.