Baffling problems under the microscope
 
Notifications
Clear all

Baffling problems under the microscope  

Page 2 / 2
  RSS
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

I have taken the extruder apart and inserted a new length of PTFE tube, with carefully adjusted ends.

The problem is at least markedly improved, possibly completely fixed. I'm not entirely certain, but it's looking promising. 😀

Publié : 20/07/2017 2:21 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

It really looks like the new PTFE tube has done the trick.

I have printed several fins without any weak-blob-lines since I have replaced the PTFE tube.

Publié : 27/07/2017 1:27 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Heureka, the problem is back and I have finally figured out the explanation.

Weak blob lines again, wrecking large prints, preferably late in the process...grrrrrr.

So here is the explanation why the problem was so baffling, and why so many different approaches to the problem are being recommended.

The underlying problem is a partial cable break in the extruder motor cables.

This causes increased resistance in the cable, resulting in voltage drop at the stepper motor.
The amount of voltage drop is dependent on the extruder position, so that the weak blob lines are more likely at certain z heights than at others. The distribution of the weak lines is not random, they appear much more frequently at the same z-levels, but not exactly at the same levels every time.

The reason for the improvement experienced when using a variety of 'fixes' (e.g. when renewing the PTFE tube, or doing cold pulls to clean the nozzle, or to somehow else reduce the friction of the filament feed system) is this: The stepper motor still receives a reduced voltage, and it either moves by one step if it has enough power, or it does not move at all if it cannot overcome the mechanical resistance.
.
So, minimizing the mechanical resistance will make the problem disappear for a while.
But eventually, the cable break worsens and the problem re-appears.
.
This also explains why I have never seen this weak blob layer problem when using BioFila filaments:
I was printing them at higher than recommended temperature, reducing the force needed for the extruder motor to push it into the hotend. And it is nice and slippery in the long PTFE tube I use to feed it to the extruder. And all the spools were new and near full, increasing the coil radius and so reducing the friction in the long PTFE tube.
.
When I installed the cable saving upgrade a few months ago, fan cables and PINDA probe cables had almost completely broken, but the stepper motor cables were apparently OK. The installation of the cable saver made it more difficult to diagnose the problem, because it slowed down it's progression.

Publié : 14/10/2017 3:01 am
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

I replaced the presumably faulty extruder motor cables, but the problem remains. Still semi-random faulty layers appear for no obvious reason, in similar but not identical locations, sometimes in clusters up to maybe 15 layers, sometimes single lines.

I'm at my wit's end.

Maybe it's intermittent failure of the heatbreak fan, but I should have noticed the fluctuation in noise levels, I think.

I had already made video recordings of display and printer earlier on, and no temperature fluctuations were displayed while the problematic layers occur.

I believe the failed layers (across repeated prints) might be distributed in a Gaussian Normal Distribution around their most likely z-level.

Could it be a failure of z-hop returning to the correct level?

I regret having thrown out multiple failed prints, because some of them had multiple failed layers but still enough cohesion of the entire object to measure it's actual z-height. Maybe that would show the cumulative effect of multiple false layer additions.

Is it possible that the printer just skipps a layer and prints 0.3mm too high? If so, then it does not show when I open the file with prusacontrol_0.9.3.390. No funny missing layers are showing in prusacontrol gcode imports where they appear in the actual prints. And normal layer progression in videos of prints where the problem occurred.....
.
Could it be an issue of a bad SD card sector or something similar? I usually delete the file from the SD card after printing, or move it to a folder on the SD card for prints I might want to repeat, and then transfer the next file to SD card. They are around 16MB in size at the moment (or 34MB when printing at 0.15mm instead of 0.3mm layer height).
Would the files end up on the same spot of the SD card and lead to similar errors.

Clutching at straws here.

Any ideas how I could progress with this?

I'll post some more pictures later on.

Publié : 16/10/2017 10:08 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Today I examined the .gcode file on the actual SD card by loading it into prusacontrol-0.9.3 , but it looks fine.

Then I prepared a .gcode file that only prints the part of the fin that is z99 and upwards. It printed fine. No weak blob lines when it's printed just above z0 level, but the last few prints always had large clusters of weak-blob-lines in that area.
Screenshot shows fin and fin-tab-tip dislodged by weak-blob-lines , and above it the nicely finished fin-tab-tip when printed by itself.

Publié : 17/10/2017 1:26 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

A different fin still on the bed in printing position, with particularly badly affected fin-tab-tip.

10h17min to print this, and it mostly fails after 9 hrs or so.

Publié : 17/10/2017 1:28 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

10 x magnification of the layers in the dislodged fin-tab-tip where the printing commenced to work as it should again:
The other screenshot shows the weak-blob lines in the middle of the same fin.

Publié : 17/10/2017 1:31 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Got up at 0400hrs to watch what happens in the last stages of the same repeat fin print.
It already had 3 weak blob lines in the usual places, so I figured it will happen at the main problem zone for sure and I'd be able to observe.
Well, it printed the fin-tab-tip perfectly but the print is useless because of the earlier faulty layers.
However, I think (again) that I figured out what goes wrong.
It's been mentioned before in one of the related threads but I cannot find it using my phone.
I think it may be the stepper motor overheating.
It was not the new PTFE tube that fixed the problem, it was a one month holiday and Winter that did the trick. Now that it's getting warmer again, the weak blob layers make a comeback.
Today I opened the enclosure during the last 2 hrs of the print, and promptly no further malfunction happened.
There may be some temperature stratification in the enclosure, coupled with varying loads for the extruder, causing the clustering of weak blob layers at certain z- heights.
.
What happens when a stepper motor overheats? Of course it will stop working, but does this lead to further increase in heat production or does it cool down again when it's not turning/stepping?

Publié : 18/10/2017 2:28 am
rotarypower101
(@rotarypower101)
Estimable Member
Re: Baffling problems under the microscope

Just wanted to say good job, keep up the effort and search.

If you suspect a issue, devise a inexpensive way to test the hypothesis.

If you think the stepper is overheating, measure the temperature in a way that suits your setup. A non contact laser would be fantastic, if you dont have it, use a contact, or even a best guess skin contact if you have to.

Then, if you suspect it is truly the issue, devise a way to cool the stepper.

Water is a fantastic heat sink in a non spillable conductive container, it would go a long way to keeping tempertures to a moderate low temperature.

Add a inexpensive adhesive copper/aluminum heat sink with thermally conductive tape. it will still help keep the motor closer to ambient. Even a small fan that could jumper off existing supply power.

Are you monitoring the temperature around the printer? Do you see a spike in temperatures of ambient?

If it is temperature related you should be able to monitor and correlate a temperature to the poor layers.

Publié : 18/10/2017 3:47 am
rotarypower101
(@rotarypower101)
Estimable Member
Re: Baffling problems under the microscope

Until you capture the bad layer in the act and have empirical data infront of you from observation, its really anyones guess.

Can you set up a video recording the process with temperature in view? assuming you are unable to sit there through the entire process until you get a bad layer.

They make some very small light standalone temperature probes for things like RC cars. might be an idea?

Have you tried a different machine if one is avalible?
Sell this one and get a replacement if there is no way to pin down the issue in your current circumstances after this much time.

Publié : 18/10/2017 4:00 am
DJ
 DJ
(@dj-6)
Trusted Member
Re: Baffling problems under the microscope

I've had various issues with what I might best call "intermittent interrupted extruder flow." I tried cleaning the nozzle; replacing the nozzle and heat break and more. Nothing helped. Finally, I thought that these damn intermittents "feel" like a software bug. What the heck, I did a factory reset after loading the latest SW (3.0.12), followed by XYZ calibration and a series of live Z adjust 75mm squares. The intermittents are greatly reduced! I was having trouble getting the gcode version of the tree frog to finish printing before (0 for 4). Now I'm 2 for 6. Can't say if this is "the fix" - just one more data point and something new to try. Good luck.

DJ

DJ

Publié : 18/10/2017 4:48 am
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Thanks rotary power, I have a new MK3 kit on order, but I' m stubborn. I'll figure this out.
.
Ambient temp within the closed enclosure was 31C this morning.
IR extruder temperature 1.5hrs after opening the enclosure was 36C, very similar to the other stepper motors, and ambient temp with enclosure open was 21C.
.
Problem is, when I open the enclosure to measure IR temp, then the fin warps and ends up with a layer shift.
.
I've been planning to install hand access holes for some time, to change filament without temp shock to the printed part.
.
Do steppers actually stop working temporarily when they get too hot, or does that just shorten their life span?
.
I thought I might go the opposite way and wrap the extruder stepper in something insulating. Then the faulty lines should appear much more frequently, or constantly.
.
I print with heat bed 80C first layer and 70C the rest, I have seen temperatures close to 40C in the enclosure in Summer (about 28C in the room with the printer).
.
Switching to low power / quiet mode might also fix my problem.
.
I made video recordings a few months ago and nothing unusual was visible or audible during the creation of the weak blob lines. Temp display stable, no garbled LCD display, no clicking or other unusual noises. But, recording quality is nowhere near as good as direct observation.

Publié : 18/10/2017 4:51 am
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope


I've had various issues with what I might best call "intermittent interrupted extruder flow." I tried cleaning the nozzle; replacing the nozzle and heat break and more. Nothing helped. Finally, I thought that these damn intermittents "feel" like a software bug. What the heck, I did a factory reset after loading the latest SW (3.0.12), followed by XYZ calibration and a series of live Z adjust 75mm squares. The intermittents are greatly reduced! I was having trouble getting the gcode version of the tree frog to finish printing before (0 for 4). Now I'm 2 for 6. Can't say if this is "the fix" - just one more data point and something new to try. Good luck.

DJ

Thanks, I was almost ready to do that yesterday and downloaded the newest and the old firmware versions.
I'm still running V_3_0_9 .
What stopped me is that I would need to boot into Windows.
And I have many near perfect prints with the current firmware, which is why I have not updated it.
If it was a software issue, then your problem should be cured entirely after the update, but it is not.
So the search continues.....
.
I unpacked various heat sinks and fans left over from previous projects, and k-type temp sensors and double sided heat sink tape, but then decided to go the easy way and try the same print again with 'Silent mode' setting. That should keep the steppers cooler.
I had only turned on the 'High Power' mode almost a year ago because I had problems with Polymaker PC-Max depositing heavy blobs of material on top of the print, and I thought it would help the nozzle to plough through these blobs without a layer shift. The solution for that problem was actually the use of the silicone sock, but complacently I left the power setting on 'High Power' ever since, although the 'Silent' mode had never actually caused any trouble (as far as I can tell).
.
I'll know in 10h 10min if this fixes the problem.....

Publié : 18/10/2017 10:12 am
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Printing in 'Silent mode' did not seem to work. Faint weak blob lines appeared early on.

So back to the basics:

Idler spring tension might be too tight, I adjusted them after some filament feeding mishap about a month ago, and then printed a few things other than fins. But the weak blob lines are back since then, as the records show. My records show:
13.7mm loaded before, 13.0mm unloaded after, 12.6mm loaded after change. That alone could be the reason.

But wait, there is more!
.
A cold pull at 150C using Polymaker PC-Max looks like the PTFE tube may no be in properly, and maybe it's sticking out further at the hobbed pulley end than before.
.
How long should a well tapered and chamfered PTFE tube in the heatbreak last? Should I schedule regular replacements?

Publié : 18/10/2017 12:36 pm
PJR
 PJR
(@pjr)
Antient Member Moderator
Re: Baffling problems under the microscope


How long should a well tapered and chamfered PTFE tube in the heatbreak last? Should I schedule regular replacements?

So many variables; impossible to answer.

It's a cheap fix and very easy to do (but easier to get it wrong!). If in doubt, swap it out.

Peter

Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…

Publié : 18/10/2017 12:40 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

But the very deformed area on that cold pull may just be an artifact, because the first cold pull looked better, although the part in the actual hot end appeared to have come off:
.
So I will not take the extruder apart just jet, but try another print overnight, with the adjusted idler tensioners.

Publié : 18/10/2017 12:43 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope



How long should a well tapered and chamfered PTFE tube in the heatbreak last? Should I schedule regular replacements?

So many variables; impossible to answer.

It's a cheap fix and very easy to do (but easier to get it wrong!). If in doubt, swap it out.

Peter

Thanks Peter!
You are right.

It looks like the adjustment of the idler tensioning screws may have been all that needed doing. Unbelievably stupid of me, the amount of filament and time I have wasted before figuring this out (again!).
.
My fin print has made it to z78.4 without issues, passing 2 of the most consistently appearing weak-blob-line levels without problems, using the most notoriously difficult filament wrt weak blob lines, transparent PLA.
.
One day I might understand how this happens. The force pulling / pushing the filament seems so much stronger than needed most of the time, why does it fail almost predictably at certain levels?

.At least it has forced me to finally install the MK2S upgrade. Will see if it improves anything now that the weak blobs seem to be banished for the moment.
.
One of the main reasons why I put in an order for the MK3 was the improved hobbed pulley setup.

Publié : 18/10/2017 9:01 pm
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Success at last!

No weak blob lines in the entire fin, although I used the most difficult material that I have tried for this purpose: Transparent PLA. It appears to be harder than other PLA filaments and harder than the BioFila filaments.

This hardness makes it less forgiving when the idler tensioner screws are too tight: The hobbed pulley cannot deform the transparent PLA very much, so the pulley is more likely to stop and the stepper motor skips a step or many.

I still don't get though why this effect is so dependent on z-height. It happens again and again at a similar z-height, not exactly the same layers, but close by.

The only thing that was wrong was the idler tensioner spring loading.
They need to be adjusted so that the space between the washers is 14mm when no filament is inserted, and 13.25mm with filament inserted.

Publié : 20/10/2017 10:08 am
jay.b4
(@jay-b4)
Active Member
Re: Baffling problems under the microscope

where you getting any click or clunking noise from the extruded stepper while experiencing this issue? My printer has just started displaying what I think are similar symptoms as to what you have been experiencing.

Here's what's happened, I was printing perfectly since assembly a few weeks ago, no problems, the bearings on the Y axis were annoying me so I switched out to the Igus and was having difficulty here, figured this out and just turned out to be that I needed to fine tune the parallel. This took a few days mind you and during this time, my filament was left on a table at room temperature. During this process I didn't touch or play around with the extruded assembly.

I reactivated my xyz, self test etc. and everything was fine. Ran V2calibrayion and the first one or two were fine as I live adjusted Z, since attempting the 3rd and subsequently several other attempts. I get a click or clunking sound from the extruded stepper. I've released the tension of the spring screws to 14mm as suggested here and elsewhere though it is still occurring. When attempting to print the v2 calibration, I get blobs initially and then fine hairline extruded lines. Vids and pics attached to outline all of this.

It's also worth noting that if I extrude manually via settings>move axis>extrude- it's absolutely fine, I extruded 500 and not a problem. As soon as I ran the v2 calibration or any other print it started happening again and I get these blobs on the printed filament before it printing hair thin. Pic attached shows this. Afterwards I extruded 200 manually again and once more it was fine. At one I point I moved the Z axis all the way to the top and the filament snapped so I'm wondering if my filament is just goosed?

Any insight you might have from resolving Your Own issue would be very much appreciated.

Publié : 29/10/2017 2:25 am
MrMik
(@mrmik)
Honorable Member
Topic starter answered:
Re: Baffling problems under the microscope

Sounds like maybe you have accidentally adjusted Tune>Flow> to very very little.

Publié : 29/10/2017 3:31 am
Page 2 / 2
Partager :