This engineer claims 100 grams direct extruder. I want one

Yes that is true. I guess 140gr is all we can hope for in this price range. Now what gearing would a 20oz.in require? 0.4A is so little.

I wonder what happened with the two projects that were presented in the posts before.

---

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

Just received my pinion and 76t gears. I've made the 18T pinion accept the 5mm motorshaft, Waiting for the MK8 to arrive. 140g vs 100g is a huge difference, i would go for the 14HR08-0654S without a doubt. The lighter 7Ncm version would be great, if it works.. .
I thought about a 2 pane damper for absorbing ringing. more on this later, first need to focus on the lightweight extruder..

What about one of the 50g Nema 14 motors with a worm gearbox?

I guesstimate a 15 or 20 to 1 ratio would be sufficient and with .9deg steps and 4 to 8x micro stepping.

A pair of delrin gears would make it low cost and maintenance. My only concern would be how to manage backlash for retraction.

With an E3dV6, I doubt you'll get 100g but pretty close would be nice.

Backlash shouldn't be an issue as you're not interested in details of the motion of the filament during reversal, only in it's final position. Any backlash is factored into the retraction distance, and since the real work of extrusion is always done in the same direction backlash doesn't factor in. The problem with large gear ratios on the extruder is that you limit the retraction speed and acceleration that can be used. For 'easy' plastics like abs and pla it's probably not much of an issue, but for some of the more tricky ones like nylon it might be.

That's a valid concern.

I'm curious as to how much of a speed hit retraction would actually take.
Using the Flex3drive as my baseline with a 40:1 reduction one would assume that retract speed and acc values could be doubled, possibly more. I guess I'll look around and see if any specific filaments are particularly difficult to work with on the flex3drive and go from there.

Quote
JamesK
Backlash shouldn't be an issue as you're not interested in details of the motion of the filament during reversal, only in it's final position. Any backlash is factored into the retraction distance, and since the real work of extrusion is always done in the same direction backlash doesn't factor in. The problem with large gear ratios on the extruder is that you limit the retraction speed and acceleration that can be used. For 'easy' plastics like abs and pla it's probably not much of an issue, but for some of the more tricky ones like nylon it might be.

But that's not how things work in the real world, especially when you have small motions commanded or to put it another way, when the backlash amount is proportionally high compared to the motion size commanded. Do a little math on some fine layer heights with small features, and a bunch of retracts and you'll see motion sizes that get swallowed in backlash easily. It's best to mechanically eliminate as much as possible.

The intention of a lightweight extruder (for me) is to get a fast motion with high acceleration, and due to limited mass a better quality in terms of ringing. I think speed of the extruder is compromised with a too high ratio,
I'm running out of excuses for not making mine. All the parts arrive today (mk8/gears/bearings), so i guess i need to start tinkering, turn on the lathe to make a holder for the spurgea and integrate that with a 5mm aluminium shaft for the MK8.. (or make a separate shaft with a printed drive dog) I'll just go ahead and start making.

Quote
Koko76
It's best to mechanically eliminate as much as possible.

Can't argue with that, but I didn't follow the rest of your argument. The retraction distance is independent of layer height, and tuned by trial and error. If there is backlash you will use a longer retraction distance to compensate, without ever having to know what the backlash was.

Quote
to3dornottobe
integrate that with a 5mm aluminium shaft for the MK8

I'd recommend a steel shaft rather than aluminum. It can be pretty difficult to get the set screw(s) to hold on a steel flat, I suspect you would chew up an aluminum shaft very quickly.

It depends, 7075 T6 is pretty tough.

Have been following this interesting thread.

Could you get a dual drive to weigh under 60 grammes?

Edited 1 time(s). Last edit at 03/13/2016 04:04PM by Mutley3D.

How much weight would we spare with aluminum-wires for the stepper coils?

Quote
o_lampe
How much weight would we spare with aluminum-wires for the stepper coils?

LOL

Quote
to3dornottobe
It depends, 7075 T6 is pretty tough.

Ooh yeah, that's some nice stuff. I only have 6061 and whatever the dubious stuff from homedepot and lowes is.

It's hard to figure out which numbers are the best to go with(some of the other threads are illuminating my knowledge in this area but its still rather dark), the motors get expressed with various torques, add some gearing to improve, get close to what? what is the required(approx) target force needed to move 1kg roll filament through an extruder? with a little power in reserve..

also with a hundred gram extruder or 200g carriage X axis motor wouldn't need much power? more than extruder or similar?
I got my carriage & extruder to 94g, but no direct drive, so was thinking about using smallest motor for X I could get away with, with a little in reserve.

Edited 1 time(s). Last edit at 03/14/2016 06:34PM by MechaBits.

Quote
JamesK

Quote
to3dornottobe
It depends, 7075 T6 is pretty tough.

Ooh yeah, that's some nice stuff. I only have 6061 and whatever the dubious stuff from homedepot and lowes is.

Typically one of the 3xxx series, otherwise known as "gummy, useless garbage".

I ordered one of the 70g pancakes to work into my setup rather than the short 11. Still picking a gearing option.

I second MechaBits's question... how to pick a gearing ratio for a specific motor torque? What are the variables we are working with?

---

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

DC42's post here is the only quantitative guidance I've seen: [forums.reprap.org]

You can take some known reference points and work from there - so a Greg/Wade geared extruder with a ratio of about 1:5 will handle 3mm filament with a fairly ordinary nema 17 (I don't have the specs to hand, but it ought to be easy to find some working examples and lookup the torque). You would need roughly 1/3 of the torque for 1.75mm filament. If you start with a given motor of known torque you can figure out the gear ratio to get an equivalent working torque to the above, and if you start from a practical gearing ratio you can figure out the needed torque.

Yeah that was the thread that shed a little on questions I had, but it's still a little voodoo to me, especially when adding the other factors likerunning motor with lower current, microsteps, etc, and i used to be quite good at algebra & trig. Thats why I was asking for the final target torque so I could work backwards, I've seent he numbers like 5:1 & 3:1 being the ideal ratio, but obviously depends on your start point plus(/x) by the the other factors, I know it would be pretty tricky to push anything past my extruder motor due to the force of the spring & hobbed cog, I probably couldnt do it by hand, but my carriage is virtually friction free, so anything would move it...a cooI breeze...obviously not when there's a motor & belt on it, but i dont think it would take as much as the extruder to move, so instead of just buying powerful motors all round I'm looking to choose a smaller one not for cost, but for invisibility, aesthetics, weight...my little pancake could probably do it, but I'm thinking of trying the 35's 14's and 11's or the other round design [www.ebay.co.uk]
but there are no torque numbers to work from, also digressing if I went for 0.9 on the X, should I also use 0.9 degree motors on the Y.

Edited 1 time(s). Last edit at 03/14/2016 09:54PM by MechaBits.

Ah, sorry, I completely missed the bit about X & Y. No idea how people work those out other than by trial and error. I don't think there's any particular need to use the same motors on X & Y, so I think you're free to use what ever works/fits best.

Quote
JamesK
Ah, sorry, I completely missed the bit about X & Y. No idea how people work those out other than by trial and error. I don't think there's any particular need to use the same motors on X & Y, so I think you're free to use what ever works/fits best.

There are online calcs to work out motor torque required to accelerate a given mass, that is a good place to start. [orientalmotor.com] has one on it's front page. There are others that one can find with a bit of digging.

yes Ive seen the link mentioned, need to register, and there is obviously tons of data out there,
[www.benjaminwilcox.com]
but sometimes I just need it spelt out in short, perhaps there is no shortcut,
I probably can do the maths with most of the given numbers, obviously speed of weight being moved comes into it too, but I'm more interested in having one understandable base line number I can work with,
eg, if most printers could get away with a 50oz stepper, (doesnt help mixing oz n g's & leverage).
or if I knew an average carriage weight to compare is 500g is a simple start point, 50mm print speed.
Then I could possibly get away with half that for my X....it's just a rough guess, I could be wrong.
obviously direct & duals have probably more weight, and tons of other hidden factors,
but all i'm looking for just a rough idea I can scale as required and base the rest of my assumptions on.

Edited 2 time(s). Last edit at 03/14/2016 11:29PM by MechaBits.

Quote
Mutley3D

Quote
o_lampe
How much weight would we spare with aluminum-wires for the stepper coils?

LOL

There were times, where aluminum wire was a valid replacement for copper in a lot of different electrical equipment. It has a higher resistance, but a stepper that is "overvolted" 2-4 times can cope with that, I'd say.

Aluminum is 3.3 times lighter than copper, but only has a 1.6 times higher resistance. So, why not think it over?

Quote
o_lampe

Quote
Mutley3D

Quote
o_lampe
How much weight would we spare with aluminum-wires for the stepper coils?

LOL

There were times, where aluminum wire was a valid replacement for copper in a lot of different electrical equipment. It has a higher resistance, but a stepper that is "overvolted" 2-4 times can cope with that, I'd say.

Aluminum is 3.3 times lighter than copper, but only has a 1.6 times higher resistance. So, why not think it over?

My joviality was perhaps in misunderstanding of your comment, but if you took the wire out of a stepper and weighed it, how much do you think it would weigh, and then what saving would you make with aluminium wire, not to mention the cost of a stepper with aluminium wire, something pretty special id imagine. The joke (at least in my mind) was the length to be gone through to gain such a small saving of what, a few grammes, ie less than 10?

If the motor is overvolted, and the ali has a higher resistance, isnt that going to contribute more heat to an item that many already have difficulty keeping cool.

The gains to be made are not in using ali wire. Or am I looking down the wrong end of the telescope?

I believe everybody got it and it was pretty amusing in fact. It's our "defect" to always try to think out of the box and that's why the craziest ideas sometimes begin as a joke...so why diesmiss it .
It would be much more interesting to find a dimensionally stable alternative to the case of the motor, that's where another big chink of its weight goes. But the same thinking goes for that: if it's not mass produced the costs would be prohibitive.

Edited 1 time(s). Last edit at 03/15/2016 07:00AM by realthor.

---

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

I think i came to a conclusion, perhaps for all the wrong reasons, I thought nema 11 57oz fits perfectly size wise but miight be more torque than needed, even though its not much smaller than a 17, as it's quite long 350g in weight not much saving, more expensive than standard nema 17, saving possibly only 50g in weight, and also trickier to fit without a custom holder...but while I might get the 11, I decided on 2 other types to play with...
there is the round nema14 at 17oz...but at £18-25 for one... or others that look identical(but a little tatty) which may be poor clones or the real deal...but at £4 worth trying, and I also got a little nema17 for £3 with wings & cog...even if I can just get these to rotate a turntable it will be a result.

Quote
realthor
I believe everybody got it and it was pretty amusing in fact. It's our "defect" to always try to think out of the box and that's why the craziest ideas sometimes begin as a joke...so why diesmiss it .
It would be much more interesting to find a dimensionally stable alternative to the case of the motor, that's where another big chink of its weight goes. But the same thinking goes for that: if it's not mass produced the costs would be prohibitive.

Please do not misunderstand me, I am not and never will be here to mock anyones ideas. I am however a realist rather than idealist. Some ideas warrant investigation and follow up, others are best left in the stable. Experience lends the capacity to differentiate between ideal and real. The biggest gains are the ones to follow. A few grammes saved which in reality would cost thousands to implement is not a big gain, nor a feasible one to attain. I do however understand a few grammes here, and a few grammes there, can quickly add up.

Yes reducing motor-can weight would be a big benefit, but there are heavy magnets aswell, and the lighter the magnet the weaker the motor, which in turn require the use of gears which may counter any weight savings. Swings and roundabouts one might say. I am trying to provoke out of box thinking, there is not much that is "out of the box" in this thread, in my own view. This may sound harsh, but if you analyse the thread content, it is hard to dispute this. Not wanting to sound slightly controversial, it may provoke some serious thought, and perhaps spur a really clever eureka moment, and different path.

FWIW I doubt it is achievable to get much below 160 grammes (including hotend) for the type of single filament extruder being discussed in this thread - ie stepper motor carriage mounted.

I think the bigger picture is being missed (although this thread was about reverse engineering an extruder that no one has seen working based on a short lifetime dc motor).

The biggest gains (and advancement for "reprap") are going to be in achieving lightweight direct drive multi filament systems. Or is reprap simply about single filament systems? As an example, I just finished assembling a proto of a dual direct drive extruder which including hotend, all fasteners and 60% infill comes in at 101grammes.

New small hot-end that will help get closer to 100g:

[plus.google.com]

---

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

Quote
Mutley3D
Please do not misunderstand me, I am not and never will be here to mock anyones ideas. I am however a realist rather than idealist. Some ideas warrant investigation and follow up, others are best left in the stable. Experience lends the capacity to differentiate between ideal and real. The biggest gains are the ones to follow. A few grammes saved which in reality would cost thousands to implement is not a big gain, nor a feasible one to attain. I do however understand a few grammes here, and a few grammes there, can quickly add up.

Yes reducing motor-can weight would be a big benefit, but there are heavy magnets aswell, and the lighter the magnet the weaker the motor, which in turn require the use of gears which may counter any weight savings. Swings and roundabouts one might say. I am trying to provoke out of box thinking, there is not much that is "out of the box" in this thread, in my own view. This may sound harsh, but if you analyse the thread content, it is hard to dispute this. Not wanting to sound slightly controversial, it may provoke some serious thought, and perhaps spur a really clever eureka moment, and different path.

FWIW I doubt it is achievable to get much below 160 grammes (including hotend) for the type of single filament extruder being discussed in this thread - ie stepper motor carriage mounted.

I think the bigger picture is being missed (although this thread was about reverse engineering an extruder that no one has seen working based on a short lifetime dc motor).

The biggest gains (and advancement for "reprap") are going to be in achieving lightweight direct drive multi filament systems. Or is reprap simply about single filament systems? As an example, I just finished assembling a proto of a dual direct drive extruder which including hotend, all fasteners and 60% infill comes in at 101grammes.

I am not misunderstanding you I totally agree with you. This thread was started not to think out the box in search for eureka moments but to see what's available to us here and now to build a lightweight extruder. It was sparked by the quite inspiring 40min video I linked in the first post.

These being said, I believe the future lies with Delta style flying extruders, that can feed a few hotends. I believe the future of multi-extruder printheads will not be in increasing the number of them, but will be as follows:
- 1 hotend that will be capable of coloring the molten plastic in the melt chamber (because coloring makes sense to me only at the periphery, a few mm deep maybe in case of chipping/scratching)
- 1 hotend delivering soluble support material
- 1 hotend that will deal with a hybrid filament made by externally gluing segments of different types of materials (a la Mosaic but hopefully open source )

So in my opinion reprap should not be about 1 hotend/1 extruder but about at least 3 of them in a lightweight setup to allow high speeds.

Edited 1 time(s). Last edit at 03/15/2016 08:15AM by realthor.

---

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

Quote
the_digital_dentist
New small hot-end that will help get closer to 100g:

[plus.google.com]

Interesting looking thing. I've been wondering if a standard heatblock has enough thermal mass, so it will be interesting to see how a very small mass like that behaves for varying extrusion rates. It's possible that shrinking the hotend so much may make it a little finicky in use.

realthor - in part i agree, in part i differ in view.

flying extruders on a delta are going to start introducing lots of harmonics to deal with when talking two or more motors, and realistically flying extruders are not cmpatible with other machine styles.

The inspiring video at the beginning of this thread - the over-riding thing i noticed was that it was not shown printing unless i missed something. I also seem to recall noticing the belt drive/motor mount setup looked a little suspect (cantilvered and suspended?)