Dual Extruder Bowden vs. Direct Feed?

Hello. I have plans to build myself a 3D printer using the Core XY system. I originally wanted to go with direct feed. However, I wanted to have dual extrusion (I really wanted to print with PLA and PVA for support material) and am unsure whether two direct feeds (probably Wade's) on the extruder will make it too heavy. Would printing with a bowden setup improve my results? Or is sticking to two direct feeds going to be better? Any help is appreciated!

For a Core XY you definitely need to go with bowden, if not you are killing the printer primary goal of speed.

Quote
ggherbaz
For a Core XY you definitely need to go with bowden, if not you are killing the printer primary goal of speed.

The advantage of a CoreXY mechanism over a Cartesian one is that you don't have to move the Y platform, which is usually a relatively heavy subassembly.

The usual extruder for the Prusa i3 is a Greg's Wade's Geared Extruder attached directly on the X carriage, which is also heavy because of the stepper.

So in a Cartesian printer you have two separate masses, each around 500g, that you have to accelerate and decelerate in perpendicular directions, for each movement. That is a lot of kinetic energy (mv^2) that you have to generate and then dissipate. It is this amount of kinetic energy that basically determines how fast a printer can go, because any structure can only be made that much rigid. The Prusa i3 is usually set to work at round 40~50mm/s. Doubling the speed would result in four times the kinetic energy, so to keep the kinetic energy constant, one would have to diminish the moving mass by a factor of four.

With a single extruder, this is possible by using a Bowden system. The mass of the X-carriage can be brought down to around 500/4=125g. However, there is no way to reduce the mass of the Y platform by a factor of four in a Prusa i3. Consequently, there is not much speed to be gained by switching to a Bowden system on a Prusa i3.

The situation for a CoreXY is completely different, and here the Bowden system can really bring down the mass of the moving head, just like in a linear delta printer. This is why most linear delta printers and CoreXY printers use a Bowden system.

You can roughly estimate that having dual extruders will double the mass of a moving head, so it will also double the kinetic energy of the head, and will limit the maximum printing speed achievable with either a linear delta or a CoreXY implementation, compared to a single extruder.

Edited 1 time(s). Last edit at 03/23/2015 06:02AM by AndrewBCN.

Have you considered one direct and one Bowden?

Some rough thoughts about it. We know acceleration is more important than speed which is not realy a problem, because filament can't fuse instantly. We all agree a low head weight allows better accelerations. At the same time, the more torque your motors allows, the more weight you can deal with. As far the rails can stand the load, weight is not a factor if your motors are powerfull enough. BUT, the need of torque increase exponentialy with the load. Through, with the coreXY system, most of the time you benefits of two motors to move the head. BUT not always: for shapes, one single motor is running. That means two motors running does not mean the torque available is doubled. So we have two vectors we can step in : load and torque.

Bowden is a drastical way to reduce load; it's an easy way, and you will save about 0.25lbs per motor. There's other ways like using smaller motors for a direct feed extruders. A small continuous geared servo is lightweight (0.08lbs) but runs only at 80 rmp. Is it fast enough for an extruder ? All that needs calculation. Otherhand, a geared Nema 8 to 14 can reduce load too.

Otherwise, on any reprap, motors always runs at low speed and the travels are short. So we can experience vibrations and resonance. The reprap world uses sometimes geared motors for resolution purpose. Because planetary gear's reduction ratio is very high, the acceleration won't be good enough. But what about a low ratio ? I mean 1:2 or 1:3 ratio. That could reduce vibrations, avoid resonance, allow more load, and at the same time allow good accelerations. With torque increased two or three times, It should be possible to manage more load, and to reach a good speed with an acceptable acceleration. That should allow direct feed multiple extruders.

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

How about making a direct-drive extruder using a small geared DC motor with a shaft encoder? It should be much lighter than a stepper motor.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

True, but the encoder management will be very exotic.

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Quote
Zavashier
True, but the encoder management will be very exotic.

Not really, it's been done before with a small 8-bit microcontroller driving an H-bridge chip, with step and direction inputs to emulate a stepper motor.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Easy for you It sounds like I don't have enough electronics skills to manage that !

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Im biased as a delta owner but I want to see flying extruder on Prusas.

My first extruder was a small DC gear motor with a shaft encoder. The problem is small DC motors wear out relatively quickly when running for hours continuously at high torque and can't reverse very quickly due to the fast spinning rotor.

The way to go is probably a brush-less DC servo motor at the fixed end of the Bowden tube and a shaft encoder and pinch wheel at the entrance to the hot end with closed loop control.

---

[www.hydraraptor.blogspot.com]

nophead, any firmware for that setup? I'd love to implement that on my Kossel.

The standard way to handle servos is with a PID loop but it might need some trick to prevent oscillation due to the dead zone created by the slack in the tube. I might have a go at it some time towards the end of this year if somebody doesn't do it sooner.

---

[www.hydraraptor.blogspot.com]

Or use a Flex3Drive if you havent heard of it
(sorry i couldnt help myself )

Quote
nophead
The way to go is probably a brush-less DC servo motor at the fixed end of the Bowden tube and a shaft encoder and pinch wheel at the entrance to the hot end with closed loop control.

No need for a servo, just use a stepper as normal.
[www.dynomotion.com]

A servo would require dual encoders to be stable.

Edited 1 time(s). Last edit at 04/09/2015 01:12PM by 691175002.