Inconsistent extruder feed, jams - tried new hotend with no luck

Quote
cristian

Quote
geemoto
I realized that I actually have the extruder output data as well, so we'd be able to see if the extruder was simply outputting more power to compensate for the temperature swings, it doesn't seem to be the case - at least not to a huge degree

Too noisy to tell...

BTW, what nozzle diameter are you using?

0.4mm. Wiggling the heater cartridge/thermistor doesn't seem to have any noticeable effect.

Edited 4 time(s). Last edit at 06/23/2015 08:11PM by geemoto.

I printed the same thin walled object in black PLA. Now it looks like there is a single layer in which you can see light between it and the layer above it, as well asthe layer below it. If the Z went up too high on that layer, the layer above would be closer, not equally far. I'm guessing either the extruder under-extrudes that layer or there is some sort of wobble. I tried a print where I was putting some weight on the X carriage to reduce backlash - didn't seem to improve so I don't think it is backlash.

E.g.

------Normal Layer-------
------Normal Layer-------
------Normal Layer-------
------Small empty area---------
------Suspect Layer ---------
------Small empty area---------
------Normal Layer-------
------Normal Layer-------
------Normal Layer-------

I also noticed that there seems to be some inconsistencies in feed for a flat layer, this could further point to the extruder?? But it also seems to happen at the same spot in the print every time, which bring me back to thinking its something to do with the z-axis.

Edited 2 time(s). Last edit at 06/23/2015 07:57PM by geemoto.

My PID parameters would not work for you because my heater block is not standard size: the hotend I have is not very known, and this is unfortunate because this hotend is virtually impossible to jam (I have been trying for almost a year now, and I never managed). If you cannot solve your problem otherwise, you may want to have a look at it.

Meanwhile, I think you have to play with cooling, isolation and temperature to avoid jamming, since probably you don't have any specific hardware or electronic issue.

From the pictures you posted I cannot see if your heater block is thermally isolated: if it is not, this may be a reason for the PID overshooting or oscillating.

Quote
cristian
My PID parameters would not work for you because my heater block is not standard size: the hotend I have is not very known, and this is unfortunate because this hotend is virtually impossible to jam (I have been trying for almost a year now, and I never managed). If you cannot solve your problem otherwise, you may want to have a look at it.

Meanwhile, I think you have to play with cooling, isolation and temperature to avoid jamming, since probably you don't have any specific hardware or electronic issue.

From the pictures you posted I cannot see if your heater block is thermally isolated: if it is not, this may be a reason for the PID overshooting or oscillating.

I actually deleted all that stuff in the post about PID tuning, you must have caught it right before the update I was using the wrong command to send the new parameters (d'oh!), now it is quite stable. It is thermally insulated with a boot on all sides but the top [img.tapatalk.com]

I think the thermal isolation is pretty good, at 200C I an touch the first fin (closest to hotend) and it feels like only ~45C. I'm wondering if it's something to do with friction in the hotend...maybe exacerbated by filament swelling due to water absorption (I've had this filament for about 8 months now).

If there's a test I can do to determine whether its the extruder or Z-axis that would be great...I'm fairly convinced it's one of those two at this point. I'm leaning towards Z because the defects are in almost the exact same spot every time.

Edited 1 time(s). Last edit at 06/23/2015 08:55PM by geemoto.

Quote
geemoto
I'm wondering if it's something to do with friction in the hotend...maybe exacerbated by filament swelling due to water absorption (I've had this filament for about 8 months now).

Even if friction is increased by water, you should be able to print properly. Otherwise it means that your extruder is working at its limit, at least with that type of filament. Maybe you have been unlucky with your extruders, or the stepper motor(s) you are using is simply not powerful enough for the speed you want to reach.

Edit: by the way, I read many times that when the filament absorbs a lot of water, you should have bubbles in your print. I never had this problem, and from your pictures neither you do.

Edited 2 time(s). Last edit at 06/23/2015 08:51PM by cristian.

Yep, that makes sense. If I can be more confident it's the extruder and not something to do with the z-axis I could change the extruder config. But I'm worried I'm going to go through all this hassle to change it and it won't even be the problem

Another thing to exclude: is the filament unrolling smoothly from the spool?

Yet one more: are you sure you are not overextruding? Although we may rule this out anyway because with spiral vases even if you overextrude a little it is not a problem.

The filament is rolling fairly smoothly, there's some resistance. On a few other prints I unrolled a bunch of filament before the print and the problem persisted, I don't think it's that.

Not sure about overextrusion, I calibrated the extruder with no hotend, 112 steps/mm. With the hotend attached, its closer to 120steps/mm. I'm currently running at 120steps/mm

Given the extruder doesn't skip or slip, is it still possible for uneven extrusion? E.g. if the hotend gets 'jammed', and the extruder still has enough force to push, could the filament compress in the hotend, and then release the pressure all at once when the 'jam' is cleared? If that could happen, then simply getting a more powerful or geared extruder may not solve the problem.

Quote
geemoto
Given the extruder doesn't skip or slip, is it still possible for uneven extrusion? E.g. if the hotend gets 'jammed', and the extruder still has enough force to push, could the filament compress in the hotend, and then release the pressure all at once when the 'jam' is cleared? If that could happen, then simply getting a more powerful or geared extruder may not solve the problem.

If the temperature is stable and you have a good quality hotend, this effect should be negligible.

Edit: now that you make me think about it, I had this issue with an hotend with too long transition zone, that was not suitable for PLA. I also got these intermittent "almost clogs", even worse than yours. It got a bit better by setting cooling at 100% and increasing the printing temperature to 220C or more. However it was impossible to use with PLA.

Edited 1 time(s). Last edit at 06/24/2015 11:35AM by cristian.

That flat print is interesting - doesn't the intermittent thickness on a single layer rule out anything to do with Z? Your comments about the extruder skipping make me wonder. Is the grub screw for the gear on the extruder stepper motor firmly against a flat on the shaft? It might be worth pulling the gear off and looking for any witness marks for slipping, or if you can see the shaft at the end of the gear you could make a tell-tale mark with a pen and see if it's slipping during a print. You said the problem gets worse with lower layer heights, and that might fit with back pressure in the hot end causing the extruder to slip. The apparent Z relationship on the tall tower might just be because each raise in Z is a chance for the pressure to reduce and free things up. Maybe. Kind of grasping at straws

Quote
cristian
If the temperature is stable and you have a good quality hotend, this effect should be negligible.

After calibrating the PID controller, the temperature is quite stable, +/- 2C. I after 3-4 oscillations, the largest of which usually isn't more than +/- 4C. I 'think' the hexagon is supposed to be a pretty good hotend, it does have a fair length transition zone, maybe about 2.5cm. But, I doubt there is too much melting happening in the finned zone since I can touch the point closest to the hotend with my hand and its temp is only ~40C.

Quote
cristian
Edit: now that you make me think about it, I had this issue with an hotend with too long transition zone, that was not suitable for PLA. I also got these intermittent "almost clogs", even worse than yours. It got a bit better by setting cooling at 100% and increasing the printing temperature to 220C or more. However it was impossible to use with PLA.

I figured cranking the temp up to 220 or 240 would help, it seems the same problem persists. Though it seems masked because the hotter layers tend to fuse together more. If the hexagon hotend isn't good enough, I don't know which hotend would be!

Quote
JamesK
That flat print is interesting - doesn't the intermittent thickness on a single layer rule out anything to do with Z?

I would think so, but the flat layer could also be due to poor bed leveling since it is the first layer. I would agree that this makes it seem less likely it is Z.

Quote
JamesK
Your comments about the extruder skipping make me wonder. Is the grub screw for the gear on the extruder stepper motor firmly against a flat on the shaft? It might be worth pulling the gear off and looking for any witness marks for slipping, or if you can see the shaft at the end of the gear you could make a tell-tale mark with a pen and see if it's slipping during a print. You said the problem gets worse with lower layer heights, and that might fit with back pressure in the hot end causing the extruder to slip. The apparent Z relationship on the tall tower might just be because each raise in Z is a chance for the pressure to reduce and free things up. Maybe. Kind of grasping at straws

I'll look into this, there actually wasn't a grub screw at all a few days ago, I think the gear might have a flat in it already - I didn't actually build it myself and have never taken it apart. The last time I tried it wouldn't really budge, but I can give it another shot! This is definitely something that I've thought of before. I put a grub screw in pretty tightly a few days ago thinking this might fix the problem, no luck. Maybe I'll try to get it off one more time, I was worried I'd rip the stepper shaft right out of the stepper, is this a reasonable concern? I was prying it pretty good with a flat head screw driver to no avail.

That's odd, I wonder how they fixed it on there. If they used thread locker you might have to use a heat gun to loosen it up, but I've not heard of anyone using that on a stepper motor shaft. I'd be worried about damaging the bearings if you have to use too much force to lever it off. Presumably if it's that hard to remove it isn't slipping, but a tell-tale mark would check that pretty easily.

Quote
JamesK
That's odd, I wonder how they fixed it on there. If they used thread locker you might have to use a heat gun to loosen it up, but I've not heard of anyone using that on a stepper motor shaft. I'd be worried about damaging the bearings if you have to use too much force to lever it off. Presumably if it's that hard to remove it isn't slipping, but a tell-tale mark would check that pretty easily.

From the looks of it, I think what happened is they forgot to put the grub screw, and then after running it it got jammed up. The screw doesn't line up with the flat anymore, and I think that's why it doesn't want to come apart. But my thoughts exactly, if its that tight, its doubtful it's slipping.

How fast should an extruder be able to extrude without skipping? I seem to remember the extruder used to be able to do 20-30mm/s and now it struggles to do 8mm/s

I'm not sure what the upper end of extrusion speed is. It's obviously going to depend a lot on nozzle diameter and the heater power. 20-30 mm/s of filament into the extruder sounds like a lot though. I did a back of the envelope calculation for a printspeed of 100mm/s, 0.3 layers with 0.4mm width and 1.75mm diameter filament and only got 4.2 mm/s of input. Lower print speeds, thinner layers or bigger filament diameter would all reduce that. Maybe I got the numbers wrong.

Quote
geemoto
I 'think' the hexagon is supposed to be a pretty good hotend, it does have a fair length transition zone, maybe about 2.5cm.

2.5cm is long, I would not be surprised if you had problem with that hotend then. In my hotend more than 90% of the temperature transition happens in less than 1.5cm.

Quote
geemoto
If the hexagon hotend isn't good enough, I don't know which hotend would be!

The one I wrote you about would not disappoint you, it was designed specifically for PLA although it can print anything below 245C (it is not all metal). However, if others have your same hotend and can print well with it, it means they are doing something different from what you do or have a different setup.

Have you tried to print with other than PLA to see if the problem goes away?

Edited 1 time(s). Last edit at 06/24/2015 04:29PM by cristian.

Quote
cristian

Quote
geemoto
I 'think' the hexagon is supposed to be a pretty good hotend, it does have a fair length transition zone, maybe about 2.5cm.

2.5cm is long, I would not be surprised if you had problem with that hotend then. In my hotend more than 90% of the temperature transition happens in less than 1.5cm.

Quote
geemoto
If the hexagon hotend isn't good enough, I don't know which hotend would be!

The one I wrote you about would not disappoint you, it was designed specifically for PLA although it can print anything below 245C (it is not all metal). However, if others have your same hotend and can print well with it, it means they are doing something different from what you do or have a different setup.

Have you tried to print with other than PLA to see if the problem goes away?

That was probably poor wording, it has about 2.5cm zone of fins, but I'm not sure what the transition zone size is. If I were to approximate, the distance from the ceramic heater to the first fin is <1cm. The temperature feels ~45C at this point (by touch), so the temp drop is ~150C over 1 cm.

I don't have any other filaments to test, unfortunately.

Quote
geemoto
That was probably poor wording, it has about 2.5cm zone of fins, but I'm not sure what the transition zone size is. If I were to approximate, the distance from the ceramic heater to the first fin is <1cm. The temperature feels ~45C at this point (by touch), so the temp drop is ~150C over 1 cm.

This sounds OK. Then I really run out of ideas, probably a geared extruder is the way to go.

I turned off retraction and noticed some weird gaps, don't know if this provides any additional insight =/ The gaps are where each layer starts

Edited 1 time(s). Last edit at 06/24/2015 07:26PM by geemoto.

At this point I'm thinking that "possessed by daemons" might actually be the most likely explanation

Those repeating stepped gaps between the problem layers must be telling us something, but I have absolutely no idea what!

Quote
JamesK
At this point I'm thinking that "possessed by daemons" might actually be the most likely explanation

Those repeating stepped gaps between the problem layers must be telling us something, but I have absolutely no idea what!

I'm with you, I'm ready to either call the exorcist or throw this thing out the window.

This is really boggling, the stepper doesn't even jam extruding at 165C at 5mm/s. But when you extrude at 8mm/s it starts jamming up, even at 200C =/

Edited 1 time(s). Last edit at 06/24/2015 07:39PM by geemoto.

Ah, but think of the satisfaction when you work it out. You'll be able to add your own entry to the pictorial guide for all the poor souls that might run into this in the future.

Also, it isn't backlash. I added some anti-backlash pieces and no change [www.thingiverse.com]

I was able to get a good picture that seems to explain whats happening a little bit. It doesn't seem that there are ONLY missing layers, there are layers that are darker to compensate. If you look at the bottom region, that is what it should like. As you move up, you see regions that are more transparent, and then more opaque. So it seems like the overall filament flow may be okay, but it doesn't seem to be coming out evenly. I'm not too sure how that could be happening.

Edited 1 time(s). Last edit at 06/25/2015 12:10AM by geemoto.

Thanks for all the help.

I've moved the thread to here [forums.reprap.org]

There was a lot of information scattered through posts so I tried to summarize them in the other thread so it'll be easier for others to help.