New i3V owner: A few beginner questions

Ok, it's time to return to my so far not so great 3D-printing efforts. Thanks again to everyone for their advice. Let me try to go one by one with answers and then also share some findings of my own.

1. I've checked the assembly of pretty much everything, and the one problem I found was that extruder block tilted slightly to the right (side with the motor). I could even tilt the block with my hand a little, so it was definitely something to improve. After I tightened eccentric bolt, the block now seems nice and tight, although it does look a small portion of degree out of perpendicular state.

2. I'm not sure how to check belts tightness. They seem rather tight to me, and there's no skipping or anything of sorts. Y-idler belt is not rubbing the wood bracket, but X-idler belt does stick to one side and touches the wood there. I'm not sure how to adjust that.

X-idler belt rubbing the bracket
Attachment 2069

3. It seems to me that I have a problem with the heat from extruder motor. It gets so hot during the 1-2 hour print, that I literally can't hold my finger on it longer than 0.5 second. The voltage is set to 0.37v. My original gear small deformed badly (see the picture below), and that was probably one of the reasons I couldn't get proper results after calibration. But the main extruder part was also a bit blobbed out of the box, I couldn't fit the guilder on it, so I had to use a knife to take some plastic off. As I've printed extruder myself, although the print was very bad quality, it was of a correct size, and after calibration it gave me correct extrusion results.

Original deformed gear
Attachment 2067

4. But after some time (maybe 3-4 small prints), new extruder started to deform as well. Not sure, whether it's because it is related to bad print quality, heat from the motor, PLA as the chosen material or something else. Despite this deformation (see below), it does continue to produce a much better results than the original extruder. I've printed another set of extruder parts, and some smaller things, and most of those look good enough. Although visible mesh on top layer is still present at places and some side blobbing occurs.

Deformation of the new extruder
Attachment 2068

New printed knob
Attachment 2070Attachment 2071

5. I figured out that problems with PLA adhesion only occur when I reverse the fan air flow to blow air out of extruder. If airs blows into extruder, most of it also cools the print a little, and I never have any issues with prints sticking to the glass, even without using any spray or blue tape. I can't even remove finished prints from the glass unless it cools down. What also helped, is the settings that printbus uploaded. I use "brim" setting there to add a nice 2-3mm brim around every print.

6. I've checked the voltage of my ATX power block, and it is exactly 12v.

7. Humidity in the room I print in is around 40%, and since it's quite hot these days here, I turn AC on as well.

8. If my nozzle it hot and waiting, plastic continues to extrude a little for some time. Not sure if everyone has this, or maybe it's one the problems I'm having.

9. I do feel there's an issue with slicing. If I compare bottom pattern of the original extruder parts and the one I printed (plus a couple of just first layers), it is apparent, that my lines are much thicker and there are fewer of them. If it's the case it might explain some mesh issues I'm having – where top lines are expected to be thicker than they are, so they are extruded with a greater distance from one another, producing the mesh. (photos are in the next reply)

Some latest printed results with currently deformed extruder

Attachment 2072Attachment 2073Attachment 2074Attachment 2075

Here's the comparison of the original extruder bottom layer to my results.

Original guidler vs the one I printed (first layer and full part):
Attachment 2076Attachment 2077Attachment 2078

Notice that original part has only 2 thin perimeters, which leaves enough space between perimeter and the hole to add some "cross-stitching". I've tried setting prints to 2 perimeters instead of default 3, result is the same, there's never enough space between hole and perimeter to add extra material. This is even more apparent in the next example.

Close-up of the main extruder part:
Attachment 2079Attachment 2080

Again, original part has 2 very thin perimeters, which allows for 2 perimeters for both hole and overall print. In my case perimeters can't even close properly.

This is good detail. Thanks.

The extruder sitting slightly out of level is because the weight of the motor is hanging on the right side of the carriage. You need to have the eccentric tight enough that both top wheels stay fully in contact with the extrusion, but there will always be a little flex and it will always be tilted a tiny bit. As long as it stays at the same angle, you'll be fine.

The hot motor melting the extruder is exactly what I experienced. The top screw got hot enough to pull through the extruder block, just as in your photos. This has to be solved for the printer to run consistently. I think your options are to add a fan, turn down the current or replace the motor. I guess it could also be binding, but my large gear and hobbed bolt turned freely without the motor or filament and it still melted just like yours.

I had a lot of trouble with gaps in my prints early on. They looked a lot like what you're posting. Too little plastic around the sides of tabs with holes in them. Thin walls with gaps between them, etc. I don't know exactly what fixed that issue, but it gradually got better and I'm not seeing it any more. I can tell you what I changed that got me to where I am, though.

1. Reduced retract rate to 10mm/s. Reduced blobbing and other issues related to gaps when restarting.

2. Upgraded to newer versions of slic3r. It's pretty buggy from version to version. I'm a couple of revs back, but I'm running one of the 1.0.x stable versions. Be sure to turn on variable width extrusion. This helps with tapered edges and other situations where the gap to fill doesn't match the extrusion width exactly.

3. Switched to a 3mm/.4mm Hexagon hot end.

4. Lowered my first layer height to .2mm, like all the other layers. I think the wider bottom layer extrusions were because of the increased layer height. Of course, the lower layer height makes the leveling and z height adjustment more critical. I added auto bed leveling soon after this, and I will never own another printer without it.

5. Added a print cooling fan for PLA. I always get blobby, melty results without it on small parts. What I see in your photos of the knob look similar.

Quote:

---

Originally Posted by clough42

The extruder sitting slightly out of level is because the weight of the motor is hanging on the right side of the carriage. You need to have the eccentric tight enough that both top wheels stay fully in contact with the extrusion, but there will always be a little flex and it will always be tilted a tiny bit. As long as it stays at the same angle, you'll be fine.

---

It is rather tight right now, so I believe it shouldn't affect the print quality. I will be assembling new extruder today and will try an idea I have with the motor.

Quote:

---

Originally Posted by clough42

1. Reduced retract rate to 10mm/s. Reduced blobbing and other issues related to gaps when restarting.

2. Upgraded to newer versions of slic3r. It's pretty buggy from version to version. I'm a couple of revs back, but I'm running one of the 1.0.x stable versions. Be sure to turn on variable width extrusion. This helps with tapered edges and other situations where the gap to fill doesn't match the extrusion width exactly.

3. Switched to a 3mm/.4mm Hexagon hot end.

4. Lowered my first layer height to .2mm, like all the other layers. I think the wider bottom layer extrusions were because of the increased layer height. Of course, the lower layer height makes the leveling and z height adjustment more critical. I added auto bed leveling soon after this, and I will never own another printer without it.

5. Added a print cooling fan for PLA. I always get blobby, melty results without it on small parts. What I see in your photos of the knob look similar.

---

1. Done

2. Downloaded the latest version from the beginning. It's 1.0.6, I believe. But I've also tried using Cura, which, in my opinion, produces much nicer gcode for smoother movements, but has less settings to play around with. Not sure where to find this "variable width extrusion" setting.

3. I've bought the kit with .4mm hexagon hot end.

4. Will try this, and will continue playing around with settings like that.

5. Interesting, will definitely try to print something with and without a separate fan to see if it makes a difference.

I wondered about that variable extrusion width setting too, but I think what clough was getting at is leaving the default extrusion width under print settings/advanced set to a value of zero. That tells slic3r to automatically adjust the extrusion width.

I think part of the trick with the print cooling fan is to get it as localized as possible on just the print. My playing around with a fan just pointed in the general area of the print usually led to problems keeping the top of the heat bed at temperature and/or the hot end tip at temperature. That's why people come up with various fan shroud schemes that focus the airflow. After incorporating clough42's version, my print quality definitely went up. There's a pressure related loss with the shrouds though, so you'll want a pretty strong fan on it. If the fan is connected to D9 on RAMPS, the slicers can be set to lower the fan speed if the fan turns out too strong or loud.

Personally, I'd stick with Cura. I tried it this week in part because I was frustrated with the way the newest Slic3r version (1.1.6) would spend so much time doing things "bits at a time" rather than in long passes. Once I got the first layer adjusted right, Cura along with the print cooler and retraction set to 10mm/sec started cranking out impressive quality. Well, impressive in comparison to what I was getting out of Slic3r without the cooler and retraction at the MakerFarm default speed.

Why not experiment with a lower driver setting for at least the extruder motor? As clough has pointed out in other threads, the MakerFarm setting for the drivers is questionable with the motors being used.

Just finished making a couple of mods. Installed this belt guide on both Y and X idlers. Then I assembled new extruder and before attaching the motor, I've covered the top part of it with kapton tape.

Attachment 2081

Haven't yet tested this on long prints, but I can definitely feel that top part that touches plastic is not that hot anymore. I can't touch the rest of the motor, but I can hold a finger on the taped part without any problems.


I have fan running 100% of the time, as it's connected directly to power, as per original assembly instructions. It also came with the fan shroud, which focuses airflow on the hot end cooling area, but then air moves further back, where it finds the wooden wall, and flow downwards. I think, that partially helps to cool the print as well, as I never have any adhesion problems, unless I reverse the airflow.

I'll be interested to see how the tape works out. It doesn't seem like adding insulation to the motor is the way to go. I was told by a mechanical engineer at work that the motor is designed to dump most of its heat through the machined front face. You really need to get rid of heat--not hold it in. The screw will still be conducting heat into the extruder block. I think the hot screw is primarily what causes the extrusion to yield and let it through. Maybe I'm wrong. Let us know.

I have been doing some test prints with the Makerfarm knob. I'll upload some videos when I'm done. I noticed that the design of the knob requires the printer to make a last pass around each layer, filling in a small gap at the tip of each ridge on the knob. This requires an retract/advance cycle as it moves over each gap in the knob perimeter. If your retraction is having any trouble at all, you'll get gobs of plastic built up around the sides of the knob.

Quote:

---

Originally Posted by clough42

...I noticed that the design of the knob requires the printer to make a last pass around each layer, filling in a small gap at the tip of each ridge on the knob. This requires an retract/advance cycle as it moves over each gap in the knob perimeter. If your retraction is having any trouble at all, you'll get gobs of plastic built up around the sides of the knob.

---

You're likely using a newer version of Slic3r. v0.9.9 just left the gaps. Newer versions fill in those gaps. I don't remember which version it was, but this "fix" is described in the release notes for one of the versions between 0.9.9 and 1.1.6. Something to the effect of improving fill between perimeter and infill. Generate some gcode from Cura for comparison.

EDIT: v1.0.1 and v1.1.1 mention improving gap fill

I managed to replicate the mesh top by printing on glass at 70C with no print cooling fan:

Attachment 2085

Then I re-ran the print on cold painter's tape, with a cooling fan, and the holes in the mesh are gone:

Attachment 2086

With the hot bed and no cooling fan, the first layer of bridging drooped down into the cavities in the knob, so it didn't provide any support to the top layers. Without any support, the top layers didn't squish, and they didn't spread out wide enough to touch, leaving gaps:

Attachment 2087

Otherwise, there doesn't appear to be too much difference between the two prints:

Attachment 2088Attachment 2089