Diagnosing a suddenly troublesome hotend.

[Preston_Bannister]

Diagnosing a suddenly troublesome hotend. This is a basic mk10(?) hotend, with a PTFE tube far into the heated area.

Had a long print fail about half way through, with plastic simply not extruding. Cleared the plastic out of the hotend without trouble. (But felt a bit of initial resistance pushing out the melted plastic with a metal tool?) Manually pushed plastic through the hotend w/o trouble (a few tens of mm). Started the print. Some plastic was extruded, then stopped.

Repeated the above, twice. Suspect the bit PTFE nearest the melt zone is … sticky? Marked and measured the filament pulled from the hotend, as a check.

Waiting for the hotend to cool, so I replace that bit of PTFE tube.

Make sense?

[Sven_Eric_Nielsen]

At which temperature do you usually print? What is your main filament material?

[ThantiK]

I’ve seen low quality “ptfe” degrade in this style hot end. Additionally, the ptfe generally needs to seal against the nozzle flat and if any of this screws up dimensionally these hot ends turn to crap very fast. Replace the ptfe liner.

[Preston_Bannister]

@Sven_Eric_Nielsen Depends on the print and the filament, but generally hot. This particular spool in Inland PLA+ with 205-225C suggested, printing at 225C. Did a couple of large prints prior, so this is not an immediate problem with the filament.

Do a lot of printing with “crystal” PLAs (3D Solutech, MakerGeeks, eSun) at 220-240C. (Rather prefer the translucent prints.)

[Preston_Bannister]

@ThantiK Seems you might be correct. Bit curious that the tip looks scorched, not further up (in the heater block). More heat flux through the brass nozzle? The block is aluminum, but the heat break is steel(?). Would have expected the equilibrium temperature of the PTFE tube to be more uniform. (Oxygen getting in from the tip, and reacting with the plastic tube?)

If most of the heat goes through the brass nozzle, that is a really short melt zone. (Without cutting the nozzle in half to know the inside shape.) If so then a steel nozzle on this hotend seems a bit dubious.

[Preston_Bannister]

BTW, the irony here is that I have my other (CoreXY) printer apart, to put in a better hotend. Then my i3 clone decides to quit, just when I see a need to print a better mount. Seems two printers is not enough.

Ordered an Ender3. (At $200 delivered, why not?)

[Gary_Tolley_Grogyan]

Char build up on the inside of the nozzle due to dust ingress.
Replace the nozzle.

[Preston_Bannister]

Also, the scorched end of the plastic tube is 3.85mm O.D. and the less-cooked end is 4.15mm (as is the fresh replacement cut and installed). Cut the replacement just slightly longer than the cooked original.

Hint: Test-fit the nozzle / heat block / heat break with the new liner, before putting the entire assembly together. Took substantial force to get the liner to bottom in the nozzle. Found this out the hard way.

Hint: Re-level your bed. Odds are you displaced the nozzle vertically and/or the bed, through all this. Found out the hard way.

(Neither hint is new, but recalled late.)

Printer is presently laying down a very nice raft, so much improved over prior. Will know for certain in a few hours.

[Preston_Bannister]

@Gary_Tolley_Grogyan You might be right, in part. (Hard to verify w/o sawing the nozzle in half.) However, I replaced the PTFE tube in the heated zone, and am presently printing well.

Which does not rule out char as a problem inside the nozzle. Only that is/was not the primary problem, at this moment.

[Preston_Bannister]

@Gary_Tolley_Grogyan You might be right, in part. (Hard to verify w/o sawing the nozzle in half.) However, I replaced the PTFE tube in the heated zone, and am presently printing well.

Which does not rule out char as a problem inside the nozzle. Only that is/was not the primary problem, at this moment.

[Ryan_Carlyle]

Looks to me like simple thermal degradation of the PTFE – which is something all PTFE hot end liners experience given enough time and temp – but it’s not impossible to imagine some charred PLA that crept up around the outside of the tube.

Steel is an insulator (relatively speaking), the heat break throat carries heat away, and fresh incoming filament has some cooling action on the PTFE, so yeah, it’s reasonable to expect that the PTFE inside the brass nozzle will cook a bit faster than the section inside the steel heat break.

[Preston_Bannister]

@Ryan_Carlyle Which leads to other questions.

By the Wikipedia article, pyrolysis of PTFE is detectable at 200C, but is not significant below 250C. Right. So this is pretty much the range of the bulk of FDM printing.

This is starting to sound not well studied.

One unknown is whether the original or the replacement is truly PTFE, or some cheaper blend.

After hundreds of hours printing between 220 and 240C, is this a realm where PTFE is expected to degrade, or a reflection of a substandard material?

Perhaps the PID loop and physical design of common hotends leads to the PTFE tube exposed to temperatures above 250C? Perhaps if more accurately controlled, it would be fine?

As a conservative engineer, I want to get this shit out of the hot zone, as I do not trust poorly defined and marginally catastrophic variables. (And will, in later experiments.)

[Ryan_Carlyle]

240C will eventually cause charring of genuine PTFE, yeah. Hundreds of hours sounds about right for your case. It’s a consumable part if you print anything more than 215C maybe. At 265C it lasts tens of hours, at 280 it’s hours, at 300C it’s close to instant failure. I know some people that print nylon at 265 and just change the PTFE liner after every print, but that’s definitely oddball behavior. The Micro Swiss all-metal replacement hot end is the way most people with mk10s go if they want to do high-temp printing.

If you assume the temp sensor is measuring the outside of the hot block and is not super accurate to begin with, then sure, it’s reasonable for the nozzle interior temp to be higher than nominal.

[Andrew_Blinkmann]

Is there a better liner? One that lets the PLA print at the 180 to 225 range and does not degrade over many hours?

[Sven_Eric_Nielsen]

@Preston_Bannister
It is indeed well studied and well known.
That’s the reason why I’ve asked for your printing temperatures.
The ptfe solution is neither good nor its well engineered. But this is simply these easiest and cheapest solution actually available and wide distributed on the market.
You get what you pay for.
People buy cheap printers, like you again with the ender 3 (which will show the same issue by the way). Then these people are happy for a certain period and suddenly the printer shows problems.

The solution for your problem :
Simply use quality printers or at least quality hotends with well engineered solutions.

Or to say it a little bit more direct :
Don’t blame the 3D printer community for problems which are supported by you

[funinthefalls (Anthony Bolgar)]

On a side note, have you installed the Z braces on your IIIP? They make a huge difference in print quality.

[funinthefalls (Anthony Bolgar)]

@Ryan_Carlyle I have measured a 10 degree difference from what the thermistor reads to actual temp on the MK10 blocks. I now print at 205 on all three of my IIP’s. My printrbot seems the most accurate of my 7 printers.

[Preston_Bannister]

@funinthefalls I do not believe that “Z braces” (which brace in Y) are needed, as there should not be movement or forces in Y on the XZ frame. You do not see Z braces on Prusa’s machines, and he would not fail to address, if there were such a fault.

In fact, I was immediately impressed with the Mendel90 design, when it appeared, for the elegant match of structure to forces.

Now some of the cheap i3 clones have pretty flimsy upper structures, and might benefit from “Z braces”. But that is not the case with the original Prusa, or the Wanhao/Monoprice i3 clones.

There is another case - if you have the printer on a wobbly table, then the Y movement of the bed might be moving the entire table and printer … in which case the “Z braces” might help, but … better to upgrade the table. I have also seen folk mount springy feet on their printers, which strikes me as a horrible idea. Bouncing the entire printer during the print…

I do not believe the “Z braces” are a benefit for a properly built i3 design on a solid table. Maybe I have missed something, in which case better information is welcome.

[funinthefalls (Anthony Bolgar)]

I am just going by the difference I have seen in my prints before and after the addition of braces. For the few dollars it costs to add them I found that it was worth it for me. But I print some very tall items, maybe that is why you have not seen the need for them? Just my 2 cents.

[Preston_Bannister]

@Sven_Eric_Nielsen When I bought cheap-Chinese printers, I was under no delusions as to quality. My end aim is to build a better printer (for my purpose). Yes, I hear (and believe) these mk8/mk10 extruders represent pretty much the lowest tier of the present generation. Still, they do serve well enough in a certain range, and there are a lot of folk using printers with such.

I do not recall blaming the 3D printer community for the quality of cheap Chinese printers.

I wanted units sufficient to serve as I went up the CAD/FDM learning curve, while building a better(?) printer. Got what I wanted.

There are nearly 500K(!) folk subscribed to this group. As is usual in such public forums, the fraction of folk who post is very small compared to the number reading.

The aim in writing up my misadventures is to offer well-ordered examples for other folk, as much as for my benefit. (There is a lot of dubious folklore floating around this topic.) Human nature - most folk are reluctant to ask questions, for fear of appearing foolish.

Is there a well known write-up of the usual failings of these common mk8/mk10 extruders? If so, I have missed such. Lots fragmentary…

Also, I am not overly impressed with the current generation of printers. (Though they clearly have come a very long ways from the early RepRap days.) We seem to be stuck on printers that are very solid at 30-60mm/s, with a zoo of faults when pushed faster.

Eight-bit controller software? Primitive user interfaces? Using a reactive loop to control filament temperature, via a thermally remote heater and sensor? Crude control over plastic state transitions via vague part cooling airflows? This is all going to look very primitive, in a decade.

To also be direct, my aim is to solve some of those problems. I have made and will make mistakes in public. Also gaining some ground, I believe, and documenting what I learn.