And not so hot

I’m having troubles this weekend. :frowning: Most prints fail part way through with the display showing ‘THERMAL RUNAWAY’, even though I don’t see any instance of there actually being a temperature issue. Here’s the relevant section in the log:

15:36:26.892 : X:112.77 Y:120.53 Z:0.50 E:0.00 Count X: 17237 Y:30493 Z:375
15:37:28.250 : X:99.37 Y:111.65 Z:0.50 E:0.00 Count X: 24176 Y:19099 Z:375
15:37:58.719 : X:112.98 Y:75.77 Z:0.50 E:0.00 Count X: 17247 Y:21573 Z:375
15:38:47.735 : X:89.78 Y:77.59 Z:0.50 E:0.00 Count X: 17246 Y:17297 Z:375
15:38:57.218 : X:141.29 Y:34.98 Z:0.50 E:0.00 Count X: 19808 Y:14378 Z:375
15:39:39.859 : X:153.95 Y:24.11 Z:0.50 E:0.00 Count X: 30312 Y:4231 Z:375
15:39:58.123 : X:83.66 Y:41.82 Z:0.50 E:0.00 Count X: 16338 Y:9117 Z:375
15:39:58.888 : X:85.39 Y:48.11 Z:0.50 E:0.00 Count X: 16728 Y:8378 Z:375
15:40:45.561 : X:109.82 Y:23.69 Z:0.50 E:0.00 Count X: 25640 Y:2131 Z:375
15:40:58.748 : Error:Thermal Runaway, system stopped! Heater_ID: 0
15:40:58.748 : Error:Printer halted. kill() called!

Unfortunately, Repertier-Host doesn’t put the current temperatures into the log, but I’ve attached the graphic display showing the last few minutes of the print. In this case I was running the bed at 68°C and dropping to 60°C after the first layer (I’m printing on PEI and the surface temperature is several degrees under the reported temperature). Does anyone have any idea what’s going on? Maybe a cracked wire to one of the thermistors giving instants of maximum or minimum temperature, too short to register on the display but long enough for the firmware to abort? Maybe a bug in the firmware that only shows up under my specific conditions? I am running RC8 from here…

Do you have a fan on your ramps board? It sounds like you are having a hell of a time with your mosfets. I think you actually might have over worked them trying to get over 100C. If you have a fan on your ramps keeping the mosfets and drivers cool really helps.

The firmware is doing what it is supposed to, your temps should not vary too much and when it does it shuts down for safety.

You are using my extruder and my firmware? Both your temps are all over the place, they should be extremely steady. something is fishy, do you have it near an open window or a fan or something?

Yes, your extruder and firmware. I just gave it another run with the bugfix version of RC8 and had the same shutdown, this time just about half way through the print. Come to think of it I might have replaced the heat block on the extruder at one point, so the heater and thermistor might not be the same ones I started with. I know I’ve swapped the throat (broke the first one the first time I changed the nozzle) and it’s a 50:50 chance the nozzle is the same. I’m going to swap out the hot end and wiring next, but that’s liable to be next weekend since I’m heading out of town in the morning.

No open window, sitting in the middle of the living room. It’s getting warmer outside, but the inside temp is staying pretty constant. It’s not yet warm enough outside to kick the AC in…

Is it calling run away on the bed? Try setting your bed temp and not droping it. Is your part fan hitting your hot end? Try moving it slightly out to see if your hot end temp stops jumping around.

You can disable thermal run away. But i wouldnt unless your there watching it.

I had a friend have this happen to them. Pull the thermistor off of the printer and check with an ohm meter. In his case, the cable was bad and during a print it would flex enough to cause the cable to short out mid-print. Sitting still, the cable would measure out as fine, but once printing it would randomly zero out until finally the software would kill the print.

He replaced the cable and the thermistor and all is well again.

That’s what I am leaning toward David. We’ll see next weekend if it’s the case or not.

Definitely don’t disable that safety feature. Even if it is “just” the thermistor, with the closed loop nature of the heating, a bad sensor is as much a fire hazard as a short.

It’s probably an issue with the sensor, but it might also be an issue with the heater. If the heater is failing open, then it’s not heating, even though it should be, and that would also cause a fault, although I’m not sure which fault. “Thermal Runaway” sounds like it’s getting hotter than it’s trying to get.

Also, I have “Runaway Train” stuck in my head now. It’s not entirely your fault, but I’m still going to blame you.

I looked at the code, in temperature.cpp. It looks like runaway means it’s been lower than the set temp by 4 degrees Celcius for more than 40 seconds.

      case TRStable:
        if (temperature >= tr_target_temperature[heater_index] - hysteresis_degc) {
          *timer = millis() + period_seconds * 1000UL;
        else if (PENDING(millis(), *timer)) break;
        *state = TRRunaway;
      case TRRunaway:
        _temp_error(heater_id, PSTR(MSG_T_THERMAL_RUNAWAY), PSTR(MSG_THERMAL_RUNAWAY));

I’m going to go out on a limb and say that it’s your heater that’s being inconsistent, and not your thermistor. Unless there is a lot of low pass filtering going on, I would guess that graph would be a lot crazier, like ± 100Deg if there was a wiring issue with the thermistor. Maybe there are other ways it could fail that I don’t know about…

You should be able to set your voltmeter to resistance, and check the thermistor and the heater, when they are not connected, and wiggle things a bit. Hopefully you’ll get lucky and find out which one is wrong.

It could be a lot of things, too hot or too cold, just not within spec.

Loose heater core, or faulty wiring, or faulty mosfet experiencing thermal shutdown, or a bad core.

Or just loose thermistor not making good contact, or faulty wiring, or a bad thermistor.

Near a breeze or draft, print fan facing the wrong way, bad gcode. etc.

Nah, couldn’t be any of those. I think the Norse gods are punishing me because I didn’t start my 3D odyssey with a nice troll… Instead I printed all these 3D printer parts. :wink:

I just found out my part fans were causing my printer to error out. Got a new fancy hot end with an accelerometer based auto leveling system. It sits a bit lower than my e3d hotend, so when the part fans blow, it caused the temp to drop faster than the heater cartridge can heat it back up. Then the printer thinks something is wrong so it shuts down the heaters completely. I might need to put some spacers in, having to run the fans at 20% isn’t going to cut it for some bridges. I’ve rebuilt that hot end three times this weekend trying to figure out the issue…

If it’s that same runaway temp issue, you can change the thresholds in the configuraion_adv, but the defaults are 40 seconds, which seems like a pretty long time.

OK, time for some closure… When I took the hot end off the thermistor wires just fell off, leading me to believe that might have been the source of the problem. I dropped a new hot end in and have put 20+ hours of printing in since with no problems. I haven’t yet rebuilt the old hot end (no spare thermistors … yet) but will when I get a chance. You can never have too many spares after all.

And one more thing to look for. I was seeing the same symptoms again even though the wiring was carefully routed to minimize stress. It happened several times at what seemed to be the exact same time in the print. I figured if it was that consistent it had to be something other than wiring. I had replaced the fan that blows on the print with a server fan that had higher air flow and it turns out if it ran too long it would cool the hot end enough to overcome the PID control. Changed the fan back to the low flow and the symptoms disappeared.