Remote mounted extruders.

[REdington]

When I decided to build a printer, I built the ORD Bot, but while researching printers, I have found that there are some with extruders mounted on the frame somewhere and had a tube that ran down to the hotend. The only advantage that I can see with this is that there is less mass moving on the carriage. Are there any other advantages??
I bought a wade's extruder and a hot end off of e-bay (also bought a qu-be but waiting on all the fixes) to get me up and running, but last night when I was trying my first print, a small piece that guides the filament on the top broke off and it quit feeding unless I stood there and guided it. So I going to machine a new extruder myself and mount it on the frame somewhere and try it.
I'll start a new thread on building the extruder.

Rodney

[kbob]

The term you're looking for is Bowden Tube. (Wikipedia) (RepRap.org)

As I understand it, and I'm no expert, the Bowden extruder's advantages are less moving weight, which allows faster printing speed, and a smaller print head, which may allow a bigger printable volume. The disadvantage is that it's harder to precisely control the extrusion rate when the drive motor is so far from the hot end. I think they have largely fallen out of favor.

[orcinus]

I wouldn't call them "fallen out of favour", considering some of the more famous RepRaps on the market use them (Ultimaker, Adrian's own RepRapPro Mendel, Tantillus), but they never caught on as mainstream, true.

The main issue with them is increased torque required to push the filament and more extruder backlash due to bowden tube elasticity and filament compression. This makes it harder to tune in the retraction features used to prevent oozing, mainly, but once tuned in, it shouldn't be much of an issue.

In conclusion, they're more complicated to build and harder to set up, but are necessary to achieve higher speeds (and possibly quality).

[kbob]

In conclusion, they're more complicated to build and harder to set up, but are necessary to achieve higher speeds (and possibly quality).

Thanks for clarifying. What about the Bowden would make it print with better quality?

[orcinus]

Less moving mass. Meaning less inertia, overshoot and vibration, eapecially at higher acceleration and jerk.
And on top of that, the higher the acceleration and jerk, the less blobbing and fattening you'll get at sharp direction shifts.

[cvoinescu]

Less moving mass. Meaning less inertia, overshoot and vibration, eapecially at higher acceleration and jerk.
And on top of that, the higher the acceleration and jerk, the less blobbing and fattening you'll get at sharp direction shifts.

I disagree slightly with the second half of that.

Any extruder has a non-zero response time. By that, I mean it starts extruding after the stepper starts, and the rate at which molten filament comes out of the nozzle is behind the movement of the stepper; a remote extruder much more so. Any firmware worth its salt runs the extruder at a speed proportional to the speed of the nozzle, even if that varies during a given G-code command, so having to slow down to corner, in itself, is not a cause of fattening. As I see it, the main cause of blobbing and fattening would be a slow extruder response time compared to the acceleration involved. (Jerk when starting and stopping would also cause problems, because the extruder would need to start and stop suddenly, but that is partially compensated by retraction; some jerk when cornering is good because it allows the speed to vary less.) Increasing acceleration would magnify the effect of the extruder response time: put simply, it would make it harder for the extruder to keep up with the nozzle movement. Switching from a classic extruder, with its quick response time, to a remote extruder, with its significantly longer response time, would also make the rate of extrusion depart more from the ideal correlation with the speed of nozzle movement. So I see a less responsive extruder coupled with higher acceleration as a double whammy -- definitely not an improvement. The bright side is that, with higher acceleration, more corners could be taken at high speed, so the rate of extrusion would need to vary less. In other words, higher radial acceleration available reduces the tangential acceleration necessary, and it's only the latter that applies to the extrusion rate.

However, whether the positive or the negative effects dominate, and how much of the negative you are willing to put up with to get your prints faster and/or with fewer vibration artefacts, I don't know.

[orcinus]

Even with the extruder speed exactly matching the movement speed, there are still artifacts at the line segment ends, because the extruder speed and pressure aren't linearly codependent.

Repetier FW wiki has a good article on the subject:
https://github.com/repetier/Repetier-Fi ... nt-quality

The Advance alogrithm solves that problem, but requires a lot of tuning and is disabled by default in most firmwares that support it.

Edit: I did make an error in my previous post - you don't want high acceleration. But you do want high speeds and high jerk.

[REdington]

Well I got the Wade's working for now and going to scrap the idea for now. I ordered the gears for the QU-BE that Bart is using and will change to it later.
I'll learn on this one and when I build the next one, I'll try the remote extruder and see how it does.
Thanks for all the info and links. I'll study it more before I build it.

Rodney

[gyrogearloose]

Hi, I fanally set up my Ordbot with a remote mounted extruder. I am using the rapman screw type extruder and the rapman hotend. It worked straight off without any problems. Printing speed is way faster now and the print quality with large prints is much better. The vibrations from the extruder stopping and accelerating must have affected the resolution while the extruder was still attached to the carriage. I sprayed the inside of the nylon delivery tube with silicon spray to ensure minimal friction. Dont know if that is necessary though. My next step is to design my own screw type extruder and make it more compact. I think the concept s sound and by choosing a fine thread, the motor is geared down.

[gyrogearloose]

I tried out a cool experiment with my remote extruder. I loaded the tube between the extruder and the hotend with short pieces of red and white filament. The result is a candy colored cube. I found that the predicted problems with a remote extruder are overrated. It works fine with most of the standard settings. Corners are nice and sharp, layers are consistent. The only thing I needed to change was the destring/prime rate. Prime needs to be less then destring. In my case, 1.25mm destring and 0.8mm prime. Without this, there is a lot of stringing. The reason for this, I think, is that the filament is under compression between the extruder and the hotend. The destringing needs to relieve enough of the compression force or filament will still come out. Other than that, the setup has only advantages. The nozzle is much easier to clean for example, if it gets blocked. Because the hotend is accessible from the top, you can push down a piece of filament and force the blockage out.