This engineer claims 100 grams direct extruder. I want one

I've watched 1h of the interview here with the engineer who won an innovation award in Aus for his creation, which is a 3D printer that is quite a bit different than the reprap printers around. He designed almost everything from the ground-up and, although I am very impressed by what i've seen, I am quite surprised that 100 grams extruders are achievable and I am resorting to this community to take "it apart".

I believe that is some sort of DC motor with a positional filament feedback system to make up for the non-stepping motor. And it also looks quite usual, apart its size. That much I understood
Actually looks a bit like this extruder from micron, but 3 times lighter from the 440gr og this one.

The video below is bookmarked at the extruder discussion but really is a good watch overall:

David's rPrint 3D Printer Design :: the 100gr extruder

Here's a pic of the thing:



While nothing is released in the open source realm, he claims to plan to release when the whole thing will be perfect

How close to the 100gr direct extruder can we get with what we have today? What projects are there for the lightest extruders? I am only aware of this quite recent RepRap Universal Mini Extruder which clocks in at 330gr.

But the geared Nema11 steppers are ridiculously priced, at at least twice the price of a Nema17 so bowden looks pretty light in that light.

Edited 4 time(s). Last edit at 01/29/2016 05:12AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

At HackFFM we are currently working on a similar design. It already works and weighs about the same. The problem ist that like this one it needs a closed loop controller, so it is not a drop in replacement.
With standard parts and a NEMA 14 i got a direct extruder down to 240g
[plus.google.com]
[plus.google.com]

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Quote
Srek
At HackFFM we are currently working on a similar design. It already works and weighs about the same. The problem ist that like this one it needs a closed loop controller, so it is not a drop in replacement.
With standard parts and a NEMA 14 i got a direct extruder down to 240g
[plus.google.com]
[plus.google.com]

That's a nice achievement indeed. 240 gr is impressive and I like the fact that is non-geared, at least doesn't use those incredibly expensive geared motors (compared to their non-geared counterparts). Thanks for sharing. I have noticed threaded rod on your Z or am I mistaking?

I also remembered a video that I've seen a while ago and found it now: [www.youtube.com]. I wonder how much foes this setup weigh.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Quote
realthor
I have noticed threaded rod on your Z or am I mistaking?

Yes, it's M8 threaded rod. The machine is planned as a real RepRap with very basic parts that are easy and cheap to source. The mechanical design is nearly finished, i am in the process of going over all printed parts for optimization and some smaller adjustments. Once that is done a second prototype will be built and tested. After that the final BOM is compiled.
I still have a big problem with the firmware. while in theory Marlin supports Dual X carriage that part is effectively not working currently.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

I like it. Skimmed through your blog for a little, I also like the silicone heater block cover. I wonder what is the temperature on the top of the slilicone after a few hours of use. Do you happen to have a heat measuring gun that can be used for this?

I'm sorry to go off track with this but I am really curious. Completely covering the nozzle in silicone would help with fan funnels that can be brought much closer now that melting is not an issue like with completely exposed heater block.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

If that's a cheap, brush type motor from a toy, it won't last long in a 3D printer.

Here's a stratasys extruder without the hot-end:

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

It sure looks like one from a toy, but that stratasys one looks really well made, want one.
The hotend he's using looks like one from the web sprayed black.

Quote
realthor
I like it. Skimmed through your blog for a little, I also like the silicone heater block cover. I wonder what is the temperature on the top of the slilicone after a few hours of use. Do you happen to have a heat measuring gun that can be used for this?

I'm sorry to go off track with this but I am really curious. Completely covering the nozzle in silicone would help with fan funnels that can be brought much closer now that melting is not an issue like with completely exposed heater block.

It's your thread, push it in any direction you like
I haven't measured the temperature, but it is always in a range where i can safely touch the block for a short moment, so i would expect <50C

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Very interesting. I am surprised that people still have some interest in light-weight designs. When I produced a 13 gram J-Head hot-end, there wasn't much interest.

Quote
Srek
It's your thread, push it in any direction you like

I like searchable content as much as the next guy so fo searching purposes I usually tend to stay on subject

Quote
Srek
I haven't measured the temperature, but it is always in a range where i can safely touch the block for a short moment, so i would expect <50C

50C is a good value considering that in a closed enclosure the internal temp would be about as much (maybe even 60C).

Quote
reifsnyderb
Very interesting. I am surprised that people still have some interest in light-weight designs. When I produced a 13 gram J-Head hot-end, there wasn't much interest.

Speed is still one of the biggest limiting factors for mass adoption of 3D printers and unless we can print a medium size, medium complexity part in a few hours we will remain enthusiasts thinkering around with this hobby. Some might have built businesses around it but that is a different situation. Did you improve your design over time or did you move over it? Maybe now would be a good time to revive it . I am trying for some time to talk about an induction pen-like hotend but my skills don't help me prototype such thing

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Quote
the_digital_dentist
If that's a cheap, brush type motor from a toy, it won't last long in a 3D printer.

Here's a stratasys extruder without the hot-end:

IMG_20160126_202759035.jpg

Looks like this Stratasys extruder is still using a brush-type motor, which will still wear out after a while.

It is a bit of a false claim regarding the extruder. I have seen the video, but not the machine printing or any plastic coming out, or did i blink and miss it?.

As for the claim of lightest smallest direct drive extruder.....he has to half his mass to 49grammes to beat the current smallest direct drive extruder that does actually work and print, that im aware of. Now, Im trying to recall who was the guy with that extruder......

I calculate that it takes less than 1w of motor output power (extruding 500mm/min with 10kgf) to extrude filament, which can be done with a pretty tiny DC motor provided you've got the proper gearing. The nice thing about DC servos is that you can usually get 2x to 5x the continuous power rating out of them for short periods to overcome difficult flow situations.

For reference even relatively short NEMA 17 motors can produce 3-5w of output power.

Edited 1 time(s). Last edit at 01/29/2016 03:52PM by LoboCNC.

I did a preliminary investigation into designing this type of extruder last year. A geared BLDC motor can provide much more power per unit mass than a stepper can. However, I didn't find a source of small geared BLDC motors with encoder at a low enough price. I didn't have time to look very hard, and I may pick up this project again later this year. I would make a microcontroller board to drive it so that the main printer electronics only needs to supply standard step and direction signals.

---

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].

Correct, we determined as well that a small geared dc motor is well capable to extrude filament. I was actually surprised by the force our prototype produced, it is way beyond what is needed.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

From the picture it seems the encoder is mounted in the idler gear?! Which, IMHO is the best place to really read the movement w/o gearplay.
David, maybe you don't have to look for a geared DC with encoder...

The Stratasys motor could be a brushless inrunner like these.
At least I see 6 wires coming out of the housing.
Those 3 phase motors could also be driven as steppers, like in camera gimbals. They often use the ST L6234D chip to step a BLDC motor.
I doubt they have much more torque than a bipolar stepper, but this could be an option for an open loop stepstick replacement.
-Olaf

Edited 1 time(s). Last edit at 01/30/2016 03:23AM by o_lampe.

Wonder if this motors any use?

[www.ebay.co.uk]
[www.ebay.co.uk]
they have others, doesn't look as slim, I've no idea on weight, or strength or electronics in general,
though doesnt look like it saves any size compared to a nema17, pancake might be same weight,
but I havent seen anyone using them either, if your roll of filament is on bearings surely that makes a big difference.

Edited 1 time(s). Last edit at 01/30/2016 07:37AM by MechaBits.

Quote
realthor

Quote
reifsnyderb
Very interesting. I am surprised that people still have some interest in light-weight designs. When I produced a 13 gram J-Head hot-end, there wasn't much interest.

Speed is still one of the biggest limiting factors for mass adoption of 3D printers and unless we can print a medium size, medium complexity part in a few hours we will remain enthusiasts thinkering around with this hobby. Some might have built businesses around it but that is a different situation. Did you improve your design over time or did you move over it? Maybe now would be a good time to revive it . I am trying for some time to talk about an induction pen-like hotend but my skills don't help me prototype such thing

For a long while, I though that speed was important as well. About 4 years ago, the buddaschnozzle was printing claimed speeds of near 200 mm/sec. J-Heads have been able to print at speeds of 150 mm/sec. since then. People now appear to be happy printing at 50 mm/sec.

I never revived the super-light J-Head as there didn't appear to be a demand. The entire J-Head design was greatly improved since then. Heater cartridges, dual axial thermistors, dual internal seals (to completely stop leakage), built-in properly installed bowden fittings, etc. It's all been developed. By changing materials, temperatures in excess of 280 degrees are possible. (A high temperature J-head, unfortunately, would have to have a retail cost of at least $175.) I don't think there would be much of a demand for such a J-Head anymore. So, it was never brought to market.

I think even people who mostly print Yoda heads and tugboats have decided that they'd rather have the print come out looking like the STL file than have it finish 2-3X faster. I've seen plenty of printers running at 200mm/sec and it looks impressive, but the prints they produce at those speeds do not. I design and print a lot of parts that require some mechanical precision. You don't get that at 200 mm/sec, and usually don't get it at 100 mm/sec.

The stratasys extruder uses a very small motor with a gear reduction mechanism about the same size as the motor. You won't get enough torque to move filament from a motor that size without the gear reduction mechanism.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

13g hotend sounds great, revive it, but sounds like it could be copied very quickly, add a great light tidy feeder, try to keep cost's below this one [www.ebay.co.uk]
or the new makerbot one, trying not to get copied much harder to achieve, knowing that could be the possible outcome doesn't encourage anyone.

Quote
MechaBits
try to keep cost's below this one [www.ebay.co.uk]

That's a nice design, but quite the price - close to what I spent on the entire printer. I wonder if that's how the market will go, the cheap clones at the bottom end and a niche for high-end stuff at the top. I hope it doesn't get too crazy. This part looks like solid good quality engineering, but it would be sad if we end up in the land of unobtanium and unicorn tears like the audiophile world.

Yes I've been having trouble parting with £20 for a rubber band for my belt drive turntable, the direct drive doesnt get used much either), let alone the more expensive stuff you can buy for hi end audio...when all you really need to do is get up early one morning and go foraging for nature's answer to everything.

To me that extruder looks too big for my machine, doesn't look balanced, maybe it is for a specific printer, what about 2 motors one for reel, and a smaller one with encoder for extruder? all controlled by dedicated controller(another arduino) that's what I would want for £200...

but that's not such a high price to pay if it's built to last, & if it can remove the headaches and perform flawlessly in a multitude of scenarios.

One thing many people often don't factor into the cost of a part is their time. If you buy a cheap part that is unreliable, you'll be clearing out jams and reprinting failed prints over and over, as well as wasting materials for as long as you keep trying to use that poorly designed/built part. I see this same behavior when it comes to selecting print beds and heaters, and in extruders and hot-ends, and even whole printers (the "starter" printer concept- "I'll buy something cheap now to learn about 3D printing, and get something better later").

Some people buy old, junky cars because they like to spend their time working on them. I don't understand it, but people do a lot of weird stuff. Buying a 3D printer, especially someone's first 3D printer, can be a difficult choice. While they don't understand all the technical ramifications of the design and parts selection, the price is easy to understand, so they purchase based on the price and reviews of people who have the same machine. Unfortunately, like most other things in life, if you go cheap, you usually don't get very good performance. User reviews are hard to trust because you don't know the reviewer's experience level. If someone has never used a quality, reliable printer, their review of their cheapo kit will sound like it's a good machine to an even more inexperienced reader. Unfortunately the press isn't always a big help either. The reviewers have to say good things about the product they are reviewing or they may not get more machines to review, so most reviews focus on the good stuff and gloss over the bad.

There's no guarantee that if you spend extra money you won't still end up with a poorly designed/built part that will cost you many hours of frustration. OTOH, you can be pretty sure that if you buy cheap stuff, you will be spending extra time and effort making it work. For me life is too short to spend it unclogging extruders and struggling to find a combo of materials to stack on a print bed to get prints to stick, especially when a few extra dollars get you stuff that's known to work well.

I have no experience with the Micron Tech extruder/hot-end but based on what I see, it looks like a quality product. I looked at it a year or so ago when I was selecting an extruder for my printer after having struggled with a crappy one (QUBD) and playing with my own design. At the time I couldn't find anyone who actually had one to tell me if it was as good as it looked, so I ultimately settled on the BullDog XL and E3D v6 because both had numerous favorable reviews and both appeared to be quality parts. The combo has been almost absolutely reliable. After a year of almost daily printing, I recently had my first-ever jam due to a metal particle in the hot-end that jammed the nozzle. The extruder has never ground a divot into the filament and lost grip. I spent about $200 for the two parts and the amount of time they have saved is worth 10X that amount to me.

When I look at parts to put into my printer, I compare the available choices and make my selection based on the quality and reliability they will either impart or take away from the machine and prints, then usually go with the choice that points in the direction of improvement. I occasionally make mistakes, but I learn from those and try not to repeat them. I try to use surplus industrial components where possible which often results in "over-engineering" the machine, but that sort of "error" never leads to decreased reliability or print quality so I don't mind living with it. My printer is very heavy and noisy, but I don't move it very often, and I have a basement room for it in winter and a garage for it in summer, so the noise isn't too much of a problem. Surplus industrial components, even used ones, usually perform much better, and usually cost only a little more than the stuff intended for the hobby market.

I've recently been buying linear guides via ebay for a new printer project. So far I have purchased three of them, 2 used and one new, with one bearing block each, for an average cost of <$35 each including shipping. Why use anything of lesser performance when near perfection is available so cheaply?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

the_digital_dentist:

Two important distinctions. First, some people like fixing printers, the same way some people like fixing cars. Others don't like it, but it's interesting enough that it's not worth much money to avoid it. For these people, cheap hardware may be a good idea.

Second, it seems you use your printer a lot. Many people don't, and for them it may make sense to spend some time to avoid investing hundreds of dollars in high-end parts.

All true, but maybe those folks would use their printers more it weren't such a PITA to coax a print out of them. I know that if I had to fight with the machine all the time I would quickly lose interest.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Follow-up on the Stratasys extruders:

[http://haveblue.org/?p=1600]



A lot more pictures inside and a good read overall.

I have noticed that they use gears galore but that is acceptable because the encoders provide an independent source of information. What a good thing to forget about -backlash in everything.

Now because all we have is stepper motors I am thinking of a flying extruder for a cartesian corexy printer ... that would take the weight off the rods/rails.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

realthor: You mean like [vimeo.com]

realthor: He probably means a "flying extruder" setup, relatively common on Deltas, where an extruder with a short Bowden tube is suspended loosely above the hotend using springs. The short tube reduces the downsides of a Bowden setup (elasticity, ooze / retraction) while reducing the moving mass of the effector / hotend.

The flying extruder idea was born on a Delta, but is not limited to it. A Prusa could be upgraded too. ( the moving bed is still a bottleneck)
Fixed bed printers would earn the most.
Would give it a try when using dual extruders.
-Olaf

Edited 1 time(s). Last edit at 02/02/2016 12:35AM by o_lampe.

On a Delta the flying extruder sits just above the effector with the shortest bowden tube possible to reach the circumference of the printer's bed while raising together with its 3 towers carriages, which it is fixated to -usually- by some elastic tubes. Something like this: [youtu.be]

But for a Cartesian printer is much more difficult to achieve such a thing with a short bowden tube because of its larger square (as opposed to circular) bed size that means placing the flying extruder even higher to cover the whole bed.

A better approach I am thinking of is a "print-dependent" flying extruder, when the flying extruder has an additional system to get closer to the to-be-printed area and printing a circular skirt instead or in addition to the classic skirt. This action would pull the flying extruder exactly in the center of the circle the print would be inscribed in. With this approach a very short bowden tube, for example 100mm would be possible for any size bed.

@Olaf: I don't find the moving bed being an obstacle for a flying extruder. At least not for the CoreXY system I am thinking about. The XY gantry is fixed, so the flying extruder must only comply with horizontal movement, something easily achieved by an LCD swiveling arm. They have been doing this type of assistance for ages with heavy tool manipulator arms and for a normal size bed you can basically print it all (minus screw,bearings)

Edited 4 time(s). Last edit at 02/02/2016 04:52AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff