Failing hotend thermistor?  

Re: Failing hotend thermistor?

Preheat your hotend and then maneuver your Extruder by hand while observing the temperature on the LCD, I'm certain you'll see the temp drop drastically as you're moving it around, probably all the way to the right, indicating a frayed (perhaps not fully compromised) thermistor wire.

The likely suspects are at the rear of your Extruder body, or at the entrance of the Einsy box, some pictures of those areas may help to clarify.

The temp may "drift" because the system is measuring the voltage level through a resistor divider, if the resistance changes because the cable is compromised, the computed value will fluctuate.

Hope that helps!

Re: Failing hotend thermistor?

Thanks for the hint. I'll ttry that.
Don't know if it explains the spikes as well as the drops. I can observe them on the display where the temp jerks up by 20-30 degrees as fast as it jerks down, both completely independent of x-axis position.
No way the heater cartridge could react that fast to a falsely low temp reading.

I had a broken cable on my CR-10. The graph looked like clipped above the set temperature with the spikes only going downward. You are right though concerning on what extruder position to expect the drops: on the far side of the control box.

I'll investigate over the weekend.

Cheers.

Re: Failing hotend thermistor?

Thanks for the hint. I'll ttry that.
Don't know if it explains the spikes as well as the drops. I can observe them on the display where the temp jerks up by 20-30 degrees as fast as it jerks down, both completely independent of x-axis position.
No way the heater cartridge could react that fast to a falsely low temp reading.

I see what you're saying but there's nothing between the sensor itself and the MCU reading the value but the cable and the other half of the voltage divider, it's the cable or the thermistor itself has gotten damaged.

Let us know what you discover in testing!

Re: Failing hotend thermistor?

Morning (yawns)

Tested according to the procedure you proposed. I could not find a culprit position, instead got drops erratically over the whole length of the X-Axis. Since it was a cold winter's night, I thought it appropriate to rip the printer open under the pale light of a full moon and extracted the thermistor cartridge and cabling.
Tested with an Ohmmeter and wiggled every inch of the cable without identifing one single spot of failure, but getting high resistance jerks all over the place. This was a badly faulty wiring.
Soldered a new splicing cable and put the thing back together at an hour when only the most abominable creatures walk this earth.
.
.
.
I said:"...put the thing back together..."
"...back together..."
"...back..."

I violently tore open my shirt and started running through the village shouting "PRUSAAAAA!" like a madman.

(Getting this Einsy board box closed again is a total nightmare. I forgot from assembling the machine how tedious it is to cram the cables in there and get the upper cable guide / clamp screwed together. Dammit, there would have benn room to make the top clamp a few milimeters wider. I need to find me another design or redesign it myself)

Back to topic: smooth curves again, no failure due to temp imminent.

Addendum: I got mislead by the temp jerking up, which, voltage divider or not; could only occur with a shorted-out thermistor or cabling. I didn't expect it to go above the threshold and up to the MAXTEMP range with resistance going up toward infinity. This just was not plausible. Closer inspection revealed, that in fact the cartridge was heating quite fast in order to compensate for the artificial drop in temperature. When the thermistor cable delivered correct readings again, it scared the bejeezus out of the Einsy board, which in turn pulled the emergency brake. I did not take the 40W power of the heater into account and misjudged the timescale of octoprints graphs: what looked llike jerks were actually short heating bursts. Btw, update speed on the MK3 LCD is too slow to discern minute temperature jerks while printing.

Consider that mystery solved, awaiting the next one.
Thanks for the kind replies.

Merry christmas, y'all, and a happy new year!

Re: Failing hotend thermistor?

Cheers man, glad you were able to lick it, and I think that mid-print temp jump is a-ok, I get it too and I just assume it's Slic3r magic; I haven't taken the time to investigate the g-code.

But listen, that einsy box (and all the Prusa parts) are in OpenSCAD, so you're free to make adjustments/improvements yourself, as long as you're willing to hack on some SCAD.

I loves me some OpenSCAD, but I'm nuts.

Also that clamp & textile sleeve is a big improvement over the spiral wrap and zip-tie; but I agree fixing that clamp with the Allen keys it's not fun, I've got a nice set of Wiha Hex drivers dedicated to MK3 assembly.

Happy Holidays and Happy Printing to you too!

Re: Failing hotend thermistor?

That temp jump is the start of the print. Had to re-slice because I forgot some detail, hence the waiting time for the preheated machine.
I posted this just do document nominal temp curve behaviour.

The jump is not Slic3r magic, it is lazyness. 😀 I'm too damn lazy to generate a temp preset for just about any filament in octoptint.

Being nuts is a prerequisite for 3D-Printing.

I'm a Fusion360-wimp. OpenSCAD is on the list once I'm completely locked away in an asylum 😆

Cheers!

Re: Failing hotend thermistor?

I have had both thermisters fail. My search of the forum makes me think it is not uncommon to have a break in the wire or a failed thermister itself.

Re: Failing hotend thermistor?

I have had both thermisters fail. My search of the forum makes me think it is not uncommon to have a break in the wire or a failed thermister itself.

It can happen if the wires get pinched.

The new textile sleeve harness and the Einsy box with clamps for bed and extruder harnesses are updated designs specifically to address this.

Re: Failing hotend thermistor?

I have had both thermisters fail. My search of the forum makes me think it is not uncommon to have a break in the wire or a failed thermister itself.

It can happen if the wires get pinched.

The new textile sleeve harness and the Einsy box with clamps for bed and extruder harnesses are updated designs specifically to address this.

I think I have both and it still happens. I bought a back up of both just in case.

Re: Failing hotend thermistor?

My two cents: I have spliced in a more flexible and slightly thicker copper stranded wire. My belly says it is more flexible and less prone to breaks than the original steel type of cables. My belly is not an engineer, but time will tell. I did not measure it, but I guess the difference in resistance is negligable given the relative shortness of them wires. Though "guess" and "negligable" also are no engineering level approaches...dammit, Jim, I'm a doctor, not a mechanic! 😆
Try to sort the cables a bit before wedging them into the guides of the board's case. When the whole bunch of cables gets compressed, it would help if they were not all twisted in a braid. I presume you assembled the printer from a kit? If you felt (or worse: heard) the slightest crackle when tightening the grub screw of the hotend thermistor, it is too late :geek: . One of the many things that can go wrong if one does not work in the production line at Prusa's and assembles multiple printers a day. And even with such amount of practice: humans are terribly bad concerning repeatability and following manuals. We are all musicians and have an innate urge for variation and improvisation. 😀

These cables constantly move under some amount of strain and are quite thin, so failure is a matter of time, with whatever fixation and guide. Still , the odds of both failing at the same time should be very low.
When in the printer's lifetime did you notice the error? Can you tell if it turned up suddenly of if it just worsened over time? Do you use octoprint or other software to monitor temperature changes? How do you determine that the thermistors are defective? Do they both produce "maxtemp" or "mintemp"? Do errors occur over a certain area of the printbed?

Re: Failing hotend thermistor?

I did that splicey technique with my original cable and it seem to works so I am 100% cetain it pinched and broke.