Choosing a Geared NEMA17 Extruder

I'm (finally) building myself a Mendel. I want a "latest generation" geared stepper-motor-based extruder, which seems to mean either Adrian's design, or Wade's design. Their respective wiki pages are:

• Adrian's Geared Extruder
• Wade's Geared Extruder

I saw Wade's Mendel in action at the San Mateo Maker Faire, and was impressed -- it ran fast and reliably in somewhat adverse circumstances (the only failure I recall was when someone tripped on the AC power cord and disconnected power from the Mendel!). I've never seen the Adrian-designed version.

How can I decide which extruder design is likely to be "better" for me? Has anyone got first hand experience with both extruders, who can comment on their relative merits? If not, is anyone able to compare their designs and provide "theoretical" pros and cons of each one?

In case it makes a difference: (a) I'm expecting to use PLA feedstock, and (b) being more of a software than hardware guy, would prefer an easier-to-construct design over one that needs more of a mechanical workshop (or serious mechanical knowledge) to create and repair.

If we get helpful answers in this thread, I'll happily summarize them into a wiki page.

Thanks,

Jonathan

Seems to me that the difference is what tools you have to build with. Wade's extruder needs a printer to produce it, and Adrian's could be made with hand tools and a drill press.

Since you are more a software guy, do you have the latest software to run a stepper extruder? I have built an extruder but cannot get the software to drive it.

I'm a core team member (since 2007, though I've been quiet/absent for a long time!), so I can get a full set of RP parts (direct from Adrian) including either extruder design, which is wonderful.

Assuming obtaining extruder RP parts is a non-issue, do you still see significant differences in the tools or skills needed to assemble Adrian's and Wade's geared extruders? If so, can you describe those differences, please?

I'm pretty comfortable with host software development, and I'm OK with doing some firmware debugging/fixing if that is what it takes. I'm planning to use an Arduino Mega (well, really a Seeeduino Mega, which is a clone, cheaper and with even more I/O pins) and some variant of the Pololu stepper driver electronics. I have decent (if a bit rusty) electronics skills, I can wield a soldering iron... though I'm not about to try SMT soldering!

What I *don't* want to have to do is spend a few hundred dollars creating myself a small hardware machine shop in my garage, just in order to build a working Mendel. So I do not have (and do not intend to purchase) a lathe, or a drill press, etc. I may be able to persuade local friends to let me use their garages for such things, if that becomes 100% necessary. Hand tools, no problem... I have a few, and will buy or borrow others as needed.

I'm outlining my plans and progress on my user page in the Reprap wiki, at http://reprap.org/wiki/User:Jmarsden if that helps clarify my situation.

Quote
naldarn
I have built an extruder but cannot get the software to drive it.

Do you mean you can't obtain the appropriate firmware, or that you have the correct firmware in place, but for some reason it doesn't behave as expected on your electronics? Exactly which firmware, obtained from where, are you working with? And how *does* that firmware behave for you? I suggest that you start a new topic thread about this, so we don't mix your issue with mine.

Thanks,

Jonathan

I would go with Wade's, since the printed parts make building it easier. It also doesn't take much in special tools.

I have a Wade's. The most difficult step in the construction was hobbing the drive screw, which requires a vice, power drill, file and a M3 tap.
Adrian's extruder instead uses a M4 insert with teeth, which you just drop in and know you get the proper grip. And Adrian's fits better with the x-carriage and constrains the axis less, if I understand it correctly.

I have run Wade's for some time and it is reliable and performs well. I haven't tried Adrian's yet but it has the following advantages on paper: -

Quote

It fits the Mendel better, Wade's limits the z-travel a lot.

It has a bigger gear reduction ratio so will give more torque and it will be more accurate.

You don't have to hob the drive pulley, you just buy one. Not sure how much grip it will give though.

On the other hand: -

You have to glue the insulator into Adrian's so it will be hard to replace this part that can easily fail.

Wade's has a sprung idler pulley but you could add springs to Adrian's as well if you fit longer bolts.

There isn't a set screw version of Adrian's motor gear so you have to file flats on the motor shaft. I can't make one because it was designed in AOI.

Wade's is more difficult to print because it is a shape that is very prone to warping.

Edited 2 time(s). Last edit at 05/30/2010 05:07AM by nophead.

---

[www.hydraraptor.blogspot.com]

I have Adrians Geared Extruder - and the design is OK.

The gear reduction introduces it's own set of problems though. It grips the filament fine, but the force it develops is enough to strip the threads where the brass barrel screws into the PTFE.

I tried using a Heli-coil - but it does the same thing.

I'm now on to version 4 of the hot end - for which I have high hopes.

It essentially has a stainless barrel with very thin walls as it's a poor conductor of heat. The thin walls limit the transfer of heat to the PTFE and I'm using a much courser thread to bite into the PTFE.

Don't know what Wades Extruder has to offer - but I suspect it may suffer from the same thing. Adrian's will be fine when I get this licked!

Si

I'm also trying to decide between the Wade and Adrian extruders for my first Mendel. Based on what I've read so far I'm leaning toward Wade's, but I have a couple more questions:

1. Wade's design reduces the z-travel. How much reduction and how does this compare to Adrian's?

2. Wade's wiki says that you attach the PTFE barrel by "using the two M bolts to pin it in place". I couldn't find any photos of this attachment. What are the details of this attachment?

Mahalo,
Terry

To add the PTFE barrel to Wade's extruder you insert it into it's position and then use a drill with a ~3mm drill bit to drill the holes through the PTFE barrel. Then add two M3 screws with nuts. The heads of the screws should be on the side with the gear.

Edited 1 time(s). Last edit at 07/08/2010 10:25AM by rhmorrison.

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

You see the mentioned holes in this picture from thingiverse.

rhmorrison Wrote:
-------------------------------------------------------
> To add the PTFE barrel to Wade's extruder you
> insert it into it's position and then use a drill
> with a ~4mm drill bit to drill the holes through
> the PTFE barrel. Then add two M4 screws with
> nuts. The heads of the screws should be on the
> side with the gear.

Guess I need to update the wiki again - I actually use M3 bolts, and a 2.2 mm drill for the PFTE - the extra friciton makes it tight, and there's no need for nuts on the bolts.

Wade

Thank you for the answers to one of my questions about the Wade and Adrian extruders, but no one's answered the other one yet so I'll repeat it here:

Wade's design reduces the z-travel. How much reduction and how does this compare to Adrian's

~Terry

I have a wades extruder made by rustyspoon, and have NO problems with it at all!!! it seems to be a very good setup...

(I just wish i could say the same for the hot-end)

as for the z travel.... I can go from 0 to 55 or 60mm with no problem... it seems to clear the frame OK! (using a standard hot-end.. unless if you use a longer hot-end)

cheers MB





edit: typo's

Edited 1 time(s). Last edit at 06/21/2010 06:34PM by Mdbaughman2.

Havent seen this anywhere, but adrian's design suffers from a series of issues:

1) there is always some kind of wobble and high play in the big gear movement possible reasons (even sinergic effects):

- the M4 thread used is long (75-80mm) and too weak to have a proper tensile strength and distance from the closest bearing is quite high;

- the bearings are not properly constrained (nor radial nor axial) and because of the thread is only M4, the angle at which nuts tense is more oblique than higher thread pitches, which results in having the bearing not properly paralel - this gives more play and wobble in big gear;

- the fact that big gear drive (m4) and its bearings are constrained (only!) with another bearing down the middle and with the filament in between - is basically a dunno how to call it - leverage or better scisor-ing effect which makes the big gear play inavoidable at some extent more play if the counter-bearing pressure increases; also, the filament shouldnt really play an active role in constraining the m4 thread and its bearings - i dont really think that is good;

2) the counter-flange plate with the counter bearing tends to bend over time (constant high force on it and i think its bigger than wade's and maybe partly some heat coming from the motor that is almost sticked to)

3) the perpendicularity of the motor axis on the big gear and then the paralelism of the m4 with motor axis are not ensured like they are ensured in wade's design (all being in same piece). So it requires quite a few things (3 things actually) to make them in proper alignement, and if by chance, one of them is misfit, then all are - e.g. if on the bigger block (left one) one reamed its holes a little out of alignment, then the M4 thread, bearings, and big gear drive will be out of alignment with motor axis, and same thing with the motor plate on the bottom plate;


Ways to fix / compensate:

- it doesnt says anywhere, but it really needs a very good quality M4 thread, and good material, like good steel or even A2 or A4 stainless to compensate for the above (i tried first a cutted piece from a m4 thread bzp of 1m, and that one was a failure, it resulted in having a huge play in big gear movement where the teeths of gear would come apart)

- making the small bearings to be as paralel to each other is harder than it sounds, as in practice the nuts (esp cheap ones) will screw tight at an angle most of the times. To close a nut normal play one can sharp-point hammer it on all sides in the middle, but again tighting in and out too many times a nut like that may damage the thread itself repeatedly;

- bearings outer sides length must be carefully matched to the small gap "in the wall" that has to keep them in (not too tight or their paralelism will be even more affected)

- i had to re-do the counter flange part in metal

- [reprap.org] seems to be aimed at fixing / improving at least some issues