What 3D printer...

[Parm]

I did a 20 hour print of a frame of some drawers to keep bits of the printer in under the heated bed. Mainly to see how good it was, I was pretty pleased with the results. This was with the new hot-end because I screwed up the last one due to under-tightening the nozzle and the replacement came with a shiny new extruder to replace the plastic one.

View attachment 245895

View attachment 245896

Then I decide to print off some simple clips to tidy up the cabling as they get in the way and could potentially get tangled up. Same filament, no changes to setting and it all went pear shaped, very poor prints. Here's a clip cover that is proving a challenge, many prints, all died in various stages. The symptoms are that filament stops being deposited and eventually the filament snaps at the extruder.

View attachment 245898

The raft at the front was before I started looking into the problem. Its very poor and looks like a feed issuse. The other rafts are after I tinkered with the extruder.

This is close but no cigar, the print on the right failed about 5 mins from the end

View attachment 245899

After reading around the subject, it seems that the most common causes are extruder issues, heat creep and other types of hot end blockages. After watching the next print I first looked at filament feeding, there was the occasional big clunk as the roll of filament turned and lifted over the obstruction. This was not the cause but easily fixed, it turned out that when they spool the filament on the reel, there is some left sticking out on the inside, trimming this improved filament feeding.

View attachment 245900

Next syptom was that a few minutes before the print failed, the extruder started clicking, this is where the rollers are exerting pressure on the filament, but the filament can't feed so eventually it slips in the roller, the click is when the feed roller tension is suddenly released. Below is the new extruder, same feed roller and motor, just aluminium not plastic body.

It only happens just before the print fails so I don't think this is the root cause, but after looking up causes the roller nearest, although on a bearing, several people have reported that loosening the bolt slightly so the roller can move a bit and line up with the feed hole helps. So I did that and increased the roller tension. This did improve the overall feed and the better raft shows that, but prints still failed.

View attachment 245901

If you leave the extruder clicking long enough, the toothed feed roller wears through the filament and it snaps. just where the filament enters the hole leading to the bowden tube. You can see the groove being worn in the filament below.

View attachment 245902

What I did notice is that I couldn't manually feed the filament and get anything out of the nozzle when the issue hits. However, its not a nozzle blockage, if I heat up the hot end, pull out the old filament, replace it, immediately the filement exused from the nozzle. So onto problem solving in the bowden tube and hot end.

First up the bowden tube needs to be pushed right up to the extruder and the hot end, no gaps between extruder body and tube, or hot end and tube. I checked both ends, no issues and it still happended. if it was heat creep, that is heat from the hot-end creeping up to the heat sink and melting the filament before it gets to the nozzle, I would have thought that this would also show up when I pulled out the filement whiel herating the hot-end and replaced it. So discounted that for now.

I tried different filament temps, the box was printed with PLA at 200c, the problem started with the same PLA at 200c. so I lowered it and raised it a few, same issues. I've tried it from 190 up to 205c and given I can feed the filemt manually through the nozzle at 185c I don't think this is a root cause.

Although I have changed the extruder, this was before I made the box and it is virtually identical to old one, just better quality materials. However I wlll calibrate the feed by measuring how much filament is fed when the printer thinks it has pushed 100mm into it. There's also some calibration about how much the extruder reverses the filement to play with. Its 26c in the office during this heat wave, that may be playing a part....

EDIT

Although I've learned a lot from this, my finger just fell on the buy button for the micro-swiss extruder/direct drive hot end. Life's too short and I want to get printing Its supposed to get rid of any heat creep issues and other feed related problems. I will have to calbrate the stepss per mm feed with this but it will also print a wider array of filament types as it can go hotter without heat creep so I can tinker with stuff that stock bits can't handle.

View attachment 245903

Cheers
Andy

Its all this faffing around thats putting me off buying a 3d printer. Im at a stage now where I just want things to work all the time with minimal dicking about

 

[brewdexta]

Did you make this thing yet? Stops you dropping stuff into the fan slots.

Something similar is on my list

Curious as to why you printed your clips vertically. I did mine flat on their backs, no raft and like the cover above, at high speed 75mm/s. Also bed temp 70 degrees (for PETG).

Most of the clips were printed on their back, this corner clip cover I forgot to lie down but there is obviously a problem so left it that way so I could diagnose it.

Your feed roller pressure looks way too high, my filament is barely even marked by the roller, yours looks squashed flat!

I had to increase the feed roller pressure to get the print of that raft right, but I think your right about the tension being high and there's another issue somewhere and this just symptomatic. It could be that the filament is being distorted so doesn't fit in the tube. When I get the issue you can't feed the filement by hand while its still hot. However if I remove the filement from the tube and replace it, then I can feed it through the nozzle by hand with pre-heat on.

I do have a filament feed ring I've printed and very little has gone off the roll so the feed angle is OK.

 

[brewdexta]

I've wound back the tenison to miminal, took the temp back to the same as the box I printed succesfully and will have another go. I've replace the filament in the tube, before, no feed by hand, after, sails through. Could be dimensional issues with the filament too, this was one of the rolls I got with the package, next I will try it with different filament. I shall crack it, or the new direct drive will arrive like the proverbial cavalry over the hill

Other things to check are feed calibration and lastly it could be heat creep, the latter sounds plausible but I'll do a bit more reading before taking the new hot end to bits. I need to do that when the micro-swiss one arrive as the heater and temp sensor are re-used.

 

[Screwdriver]

Did you swap out the bowden clamps and tube? The OEM version is pretty hopeless. I dd mine out of the box. Apparently the tube slips and allows a space between the tube and the heated nozzle. That gets jammed with plastic.

 

[ukracer]

I'm having trouble with some white PLA that isn't sticking to the glass bed properly and sometimes part of the print moves, I now have a polar bear with a club foot It only happens on prints with small bed surface areas.

I'll make sure the first layer is being squished down by checking the gap between the nozzle and the bed, but I've levelled it several times recently and use relatively thin economy printer paper so don't think its that.

So I've bought some Magigoo bed glue. I know some use glue sticks, some recommend water down PVA and we have loads of slugs in the garden, I could try rubbing the bed with one of those

Other than that I can play with higher/lower bed temperatures for the first layer etc. That will be easier once I have Octoprint installed, no rush, at the moment this is only an occasional annoyance and seems to be worse on this white PLA that came with the printer, not the loose free sample, my printer was bundled with 2kg on reels.

More squish.....I use 3D lac but my guess is it's hairspray without the perfume. But it works. However if the squish is not right it's not a miracle glue.

 

[ukracer]

I did a 20 hour print of a frame of some drawers to keep bits of the printer in under the heated bed. Mainly to see how good it was, I was pretty pleased with the results. This was with the new hot-end because I screwed up the last one due to under-tightening the nozzle and the replacement came with a shiny new extruder to replace the plastic one.

View attachment 245895

View attachment 245896

Then I decide to print off some simple clips to tidy up the cabling as they get in the way and could potentially get tangled up. Same filament, no changes to setting and it all went pear shaped, very poor prints. Here's a clip cover that is proving a challenge, many prints, all died in various stages. The symptoms are that filament stops being deposited and eventually the filament snaps at the extruder.

View attachment 245898

The raft at the front was before I started looking into the problem. Its very poor and looks like a feed issuse. The other rafts are after I tinkered with the extruder.

This is close but no cigar, the print on the right failed about 5 mins from the end

View attachment 245899

After reading around the subject, it seems that the most common causes are extruder issues, heat creep and other types of hot end blockages. After watching the next print I first looked at filament feeding, there was the occasional big clunk as the roll of filament turned and lifted over the obstruction. This was not the cause but easily fixed, it turned out that when they spool the filament on the reel, there is some left sticking out on the inside, trimming this improved filament feeding.

View attachment 245900

Next syptom was that a few minutes before the print failed, the extruder started clicking, this is where the rollers are exerting pressure on the filament, but the filament can't feed so eventually it slips in the roller, the click is when the feed roller tension is suddenly released. Below is the new extruder, same feed roller and motor, just aluminium not plastic body.

It only happens just before the print fails so I don't think this is the root cause, but after looking up causes the roller nearest, although on a bearing, several people have reported that loosening the bolt slightly so the roller can move a bit and line up with the feed hole helps. So I did that and increased the roller tension. This did improve the overall feed and the better raft shows that, but prints still failed.

View attachment 245901

If you leave the extruder clicking long enough, the toothed feed roller wears through the filament and it snaps. just where the filament enters the hole leading to the bowden tube. You can see the groove being worn in the filament below.

View attachment 245902

What I did notice is that I couldn't manually feed the filament and get anything out of the nozzle when the issue hits. However, its not a nozzle blockage, if I heat up the hot end, pull out the old filament, replace it, immediately the filement exused from the nozzle. So onto problem solving in the bowden tube and hot end.

First up the bowden tube needs to be pushed right up to the extruder and the hot end, no gaps between extruder body and tube, or hot end and tube. I checked both ends, no issues and it still happended. if it was heat creep, that is heat from the hot-end creeping up to the heat sink and melting the filament before it gets to the nozzle, I would have thought that this would also show up when I pulled out the filement whiel herating the hot-end and replaced it. So discounted that for now.

I tried different filament temps, the box was printed with PLA at 200c, the problem started with the same PLA at 200c. so I lowered it and raised it a few, same issues. I've tried it from 190 up to 205c and given I can feed the filemt manually through the nozzle at 185c I don't think this is a root cause.

Although I have changed the extruder, this was before I made the box and it is virtually identical to old one, just better quality materials. However I wlll calibrate the feed by measuring how much filament is fed when the printer thinks it has pushed 100mm into it. There's also some calibration about how much the extruder reverses the filement to play with. Its 26c in the office during this heat wave, that may be playing a part....

EDIT

Although I've learned a lot from this, my finger just fell on the buy button for the micro-swiss extruder/direct drive hot end. Life's too short and I want to get printing Its supposed to get rid of any heat creep issues and other feed related problems. I will have to calbrate the stepss per mm feed with this but it will also print a wider array of filament types as it can go hotter without heat creep so I can tinker with stuff that stock bits can't handle.

View attachment 245903

Cheers
Andy

Retraction setting......this affects things slowly and first causes a partial blockage. Then eventually a full blockage .
This is why you run a smaller retraction setting on a Bowden set up than you do with a direct drive.

 

[brewdexta]

Did you swap out the bowden clamps and tube? The OEM version is pretty hopeless. I dd mine out of the box. Apparently the tube slips and allows a space between the tube and the heated nozzle. That gets jammed with plastic.

When I swapped out the hot-end that I screwed up, it came with the new aluminium extruder and "capricorn" bowden tube. Before using it the first time I made sure the tube was in as far it would go before putting the retaining clips in.

So hopefully yes, but as I can't see inside the hot-end there is always that posssibility, I have seen the pictures of what happens when the tube isn't in all the way in, but I believe this is slow degredation and I've only had this new kit less than a week. But, its a learning curve, I may take the hot-end apart and have a butchers after caibrating the filament.

The current print looks good, so sods law dictates that it won't happen again and I'll never know the cause

 

[brewdexta]

Retraction setting......this affects things slowly and first causes a partial blockage. Then eventually a full blockage .
This is why you run a smaller retraction setting on a Bowden set up than you do with a direct drive.

Interesting, could this be something to do with the defaults in the slicer, I'm using Cura. I haven't changed any defaults on the machine or in Cura, just selected Ender 3. I shall have a look

EDIT

Just checked, retraction in Cura was set for 6.5mm and 25mm/s, recommended settings for ender 3 are 6mm and 25mm/s so not much different but something else to play with

 

[ukracer]

Have you got the STL's you printed
The box and the clips.

 

[ukracer]

Interesting, could this be something to do with the defaults in the slicer, I'm using Cura. I haven't changed any defaults on the machine or in Cura, just selected Ender 3. I shall have a look

Technically I am told no. I.e the STL does not contain retraction settings but the g.code will.
BUT I have seen two things in Cura. 1 probably user error in the things can change without you realising. I blamed the STL but was told it can't be that .
2. Cura versions have different defaults. The single number versions were the worst.
I always check the settings in the new versions and keep the old ones installed to test back to back.

 

[brewdexta]

Just opened the box stl files and checked the retraction settings. Its not something that comes up in the normal config by default, you have to add it. Both set the same, which is what you would suspect as the STL files shouldn't contain this info.

I may try reducing this to say 5mm to see what happens.....

 

[veesix]

Let us know how you get on with the Micro Swiss direct drive @brewdexta - it looks great. I've been considering one, also looking at flex3drive now (though it requires a bit more tinkering).

Edit: for anyone who hasn't seen them before, I've had good result with the Cura profiles from CHEP for the Ender 3: https://www.chepclub.com/cura-profiles.html

 

[Screwdriver]

At under £200 I hesitate to spend any money on upgrades because, no kidding, if I needed to spend more money, I'd buy a second machine.

The bottleneck in this process is not twiddling with the settings though that ages, not coming up with a design despite Fusion being such a steep learning curve but it is simply the gigantic print times. I have experimented slightly with speeds and feeds but that just tickles the print times. A second unit would cut them in half...

 

[ukracer]

Interesting, could this be something to do with the defaults in the slicer, I'm using Cura. I haven't changed any defaults on the machine or in Cura, just selected Ender 3. I shall have a look

EDIT

Just checked, retraction in Cura was set for 6.5mm and 25mm/s, recommended settings for ender 3 are 6mm and 25mm/s so not much different but something else to play with

Certainly the right direction for a clog. (Potentially)

 

[brewdexta]

Well, bloody thing is fine now I've turned it back to original settings

I have tried them twice before I tinkered so all i can think off is part of the roll of PLA was out of spec, or the machine is haunted!

I hate intermittent/unresolved faults.

 

[ukracer]

Its all this faffing around thats putting me off buying a 3d printer. Im at a stage now where I just want things to work all the time with minimal dicking about

The more I think, the more I am coming to the conclusion it's not for you yet.

 

[ukracer]

Well, bloody thing is fine now I've turned it back to original settings

I have tried them twice before I tinkered so all i can think off is part of the roll of PLA was out of spec, or the machine is haunted!

View attachment 245917

I hate intermittent/unresolved faults.

To be honest that's 3D printing. So many variables.
Oh there is such a thing as a partial clog. Which may or may not be visible or ever show up. Lol

 

[Parm]

The more I think, the more I am coming to the conclusion it's not for you yet.

I think you are right. I don't think im ever going to be at a level of computer wizardry and tinkeration to have fun with 3d printing

 

[brewdexta]

It's 20% into a 3.5hr print and looking good so far, I thought I would give it a bit of test to make sure it wasn't just luck.

EDIT - it piked out at 80% similar to before. So I have changed some retraction settings and see how it goes. The filament does have increased feed teeth marks and is compressed compared to filament that went through the old extruder, but the new extruder is already set to its minimum tension. if the retraction settings don't make a difference then I will look at extruder spring tension/roller pressure further. Interestingly most articles on Ender 3 extruder tension talk about how to increase not decrease it.

 

[brewdexta]

Retraction settings only delayed the inevitable so I went back to basics. First measured the filament, it should be 1.75mm +/- 0.03mm. The bits I measured before they went into the extruder was within limits. However after the extruder the pressure flattens the filament and in one direction the filament pushed 1.83mm and occasionally 1.86mm if the roller stays in one spot for any amount of time.

When I replaced the original plastic extruder for the shiny aluminium one, the cheap bowden tube was also replaced with higher quality "Capricorn" PTFE bowden tube. Feeding fresh filament into the tube with the extruder tension off is silky smooth. Pulling the filament out once squished by the extruder is an effort, so pushing it through by the extruder will be even harder.

From Capricorn's web site they say the tube ID is 1.9mm +/- 0.05mm and that filament with an OD greater than 1.8mm will cause issues.

So its the filament being distorted by too much pressure between the feed rollers and the new tube probably has closer tolerences exascerbating the problem. The trouble is there is only one feed roller tension adjuster and that is set to minimum. I have the old plastic extruder which is the same design and share many components, so compared them. The tooth roller dimensions are identical, so is the grooved roller bearing. But the old spring was a few mm shorter than the new one. So I've swapped that out to see if that makes a difference, but its beer O'clock so will test it in the morning. I will probably fix it just as the new direct drive arrives