Prusa Z-axis motor choice?

Zapp Automation (UK) are having a 10% discount weekend (What? You didn't get a coupon code? You probably aren't registered with them!) I want to order motors for my new Prusa build. I have 5 SOYO SY42STH47-1684B [www.zappautomation.co.uk], which are the largest (47mm body), double shaft, most powerful NEMA17 motors (holding torque 43.1 N-cm) they sell, on my repstrap. I want smaller motors on the Z-axis of this, so it can finally move to not having to be raised up on blocks, and wondered if the smaller SOYO SY42STH33-1334A [www.zappautomation.co.uk] (33mm body, holding torque 21.5 N-cm) would be suitable. The load on the Z-axis seems very light, and I have seen a lot of other Prusa's using smaller motors. Anyone care to comment if they already use smaller motors on the Z axis, and other axes too? Quick now, I need to order by midnight on Sunday night!

Yes I have two small NEMA17s on my Prusa z-axis. It takes next to no torque to spin the rods if they are decent quality and lubricated.

I don't understand your comments about raising it up on blocks. The Prusa motors are at the top? Even if you moved them to the bottom there would be plenty of room.

---

[www.hydraraptor.blogspot.com]

Yeah, I received the coupon and ordered a tiny NEMA11 to build Adrian's mini extruder plus a couple of SY42STH47-1684B for future projects.
In my view: the 1684Bs are really fine for the reprap uses, and the high power makes them most "recyclable" of all the lot.

I've been doing some plotting tests with the new build, that Edgar Poe guy exercised a lot the Z axis. The X and Y motors just get slightly warm, the Z motors stay cool, since they are turned off after each move. As Nophead remarks, good lubrication is a must.

We don't really want anything to run hotter on those PLA machines.

I've got a whole cache of low power steppers from discarded printers and hard disks, but have been repeatedly finding that for "real life" uses a bit more power and angular resolution are always welcome.

@Lanthan - thanks for the feedback. That reminds me, I'll order a NEMA11 too....

nophead Wrote:
-------------------------------------------------------
> I don't understand your comments about raising it up on blocks. The Prusa motors are at the top?
> Even if you moved them to the bottom there would be plenty of room.

Sorry, I was referring to my repstrap. I had cut my repstrap before I got the motors, which are at the bottom. Even if I had got the 1684A motors (no rear motor shaft) they would be slightly too long in the body. I guess you might have seen the image below of my repstrap in the other thread; it's 'up on blocks'!

I'm interested in having the motors on the Prusa I'm building at the bottom, and again I think the 1684Bs are too long. I want to try it so there isn't the weight of the X axis pulling down on the coupling, and so there is less mass at the top of the frame to cause oscillation (you touched on this in the other thread too, re frame stiffness); it will still be there, but at a different frequency perhaps, and a lighter cross-brace or the electronics mounting board can damp it out. I've found another reason as well; [www.thingiverse.com] are feet for the bottom of the Z-axis. With the motors at the bottom, I can make a cup for them and they will act as another pair of feet, supporting the Z-axis.

I'm also going to build the X-axis with the smooth rods vertical, like my repstrap and the version here: [translate.google.co.uk]
It might be very minor, but I think this layout more tightly constrains any rocking of the X-axis against the Z-axis, and seems to take less plastic to build, for perhaps a very minor loss of build height. Mine will also have LM8UU linear bearings on all axes. What could possibly go wrong?!

I expect people are getting sick of seeing this image...


Edited 1 time(s). Last edit at 08/21/2011 04:15AM by droftarts.

@Droftarts, about vibrations and motors on top:

There are actually some very serious cases where you want to have increased mass on top of a structure
[www.salvadori.org]
[en.wikipedia.org]
[en.wikipedia.org]
[www.youtube.com]

BTW the tuned mass damper on first page looks furiously like a Mendel successor ;-)
[www.rwdi.com]

---

I am also highly interested by the vertical smooth bars X axis solutions under independent development by Emmanuel and the Cypress guys. There is a potential for both significant simplfication and increased precision there.

> nophead Wrote:
> --------------------------------------------------
> -----
> > I don't understand your comments about raising
> it up on blocks. The Prusa motors are at the top?
> > Even if you moved them to the bottom there would
> be plenty of room.
>
> Sorry, I was referring to my repstrap. I had cut
> my repstrap before I got the motors, which are at
> the bottom. Even if I had got the 1684A motors (no
> rear motor shaft) they would be slightly too long
> in the body. I guess you might have seen the image
> below of my repstrap in the other thread; it's 'up
> on blocks'!

You confused me by talking about a new Prusa build. I think there is room for 1684A motors and couplings at the bottom if you take into account the bed height (when mounted on insulators) and the nozzle length.

>
> I'm interested in having the motors on the Prusa
> I'm building at the bottom, and again I think the
> 1684Bs are too long. I want to try it so there
> isn't the weight of the X axis pulling down on the
> coupling, and so there is less mass at the top of
> the frame to cause oscillation (you touched on
> this in the other thread too, re frame stiffness);
> it will still be there, but at a different
> frequency perhaps, and a lighter cross-brace or
> the electronics mounting board can damp it out.

Diagonal cross bracing from one corner to another would work but that clashes with the bed and extruder. Bracing part way down has little effect because the middle of a rod is easy to bend sideways. This is the fatal flaw in using a triangular prism to surround a cube build space. I am using a flat sheet for the base and another mounted at right angles for the XZ plane, with a big hole cut through it for the build volume. A sheet is very strong along its plane, even with a big hole in it.

> I've found another reason as well;
> [www.thingiverse.com] are feet
> for the bottom of the Z-axis. With the motors at
> the bottom, I can make a cup for them and they
> will act as another pair of feet, supporting the
> Z-axis.
>

They do work fine at the top and, as Lanthan points out, you want more stationary mass to oppose movement, not less. The design I am working on has them at the bottom though, so that the Z position is referenced to the bottom, where the Y axis is.


> I'm also going to build the X-axis with the smooth
> rods vertical, like my repstrap and the version
> here:
> [translate.google.co.uk]
> u&tl=en&u=http%3A%2F%2Fwww.skb-kiparis.ru%2F
> It might be very minor, but I think this layout
> more tightly constrains any rocking of the X-axis
> against the Z-axis, and seems to take less plastic
> to build, for perhaps a very minor loss of build
> height. Mine will also have LM8UU linear bearings
> on all axes. What could possibly go wrong?!

I am using LM10UU because 8mm rods sag slightly in the middle due to the weight of the extruder. This makes a flat bed appear slightly saddle shaped when you measure the distance to the nozzle. If you mount them vertically the bars will twist a little giving a Y error as well as Z.

---

[www.hydraraptor.blogspot.com]

Heading a bit off topic now, but what the hell, I think I'm going to buy the smaller motors for the Z-axes...

@nophead - sorry about the confusion, I'm actually talking about putting the smaller motors on the Z-axes of BOTH my repstrap AND new Prusa build, and both machines will have the motors at the bottom. My repstrap has 5 of the 1684B motors at present, so I can swap out two of them for the smaller motors, and use the two larger motors on the new prusa. So I'll get 4 x 1334A and one 1684 to finish the Prusa build.

I take the point of using more weight to damp vibration. Yet the original Mendel copes okay without extra mass at the top, doesn't it? That's what I had in mind originally. If it does cause problems, I have a plan; I have two 650mm lengths of threaded rod left, which I could install at the top of the frame (where the Prusa motor mounts would normally attach to). I can then triangulate onto the frame sides without getting in the way of anything. It's certainly going to be wide, but should be very rigid - I could even mount the filament spool on it. If it all doesn't work, I can always revert to motors on top!

Like you say, the only way to fully constrain the axes is in a box. I like the Makerbot/Ultimaker style because of this, but I don't have a laser cutter or router table. The other advantage of a box type printer, especially for a noted supplier of parts such as yourself, is that it is effectively its own chamber. I also like the Ultimaker X/Y axis set up, that they are not stacked on each other, and they come together at the print head/hot end; it looks very elegant, even though it uses more smooth rods. I don't see any reason why this couldn't support a direct-drive extruder, rather than use a Bowden cable. Also, the motors could be mounted on the outside, so wouldn't get as hot as inside the chamber. Hmm, another day, another project added to the list.

Good point about 8mm rod stiffness, but I think I'm going to have to keep that standard on this one. I'm using SFC8 hardened ground rail, which is pretty tough stuff. I saw the saddle shape on jcabrer's blog, and wondered momentarily what might be the cause.

> If you mount them vertically the bars will twist a little giving a Y error as well as Z.
Is your concern that the extruder drive weight is off-centre? I can see that might be a problem. Otherwise I don't see why it should be any worse than the existing horizontal rails.

I still haven't soldered that resistor on the Gen6 electronics yet... my soldering station seems to have got a bit lost when I moved house recently!

> I take the point of using more weight to damp
> vibration. Yet the original Mendel copes okay
> without extra mass at the top, doesn't it?

Not really. It is a bit better than Prusa because the vertex brackets are much more chunky and the z-brackets hold the bottom of the z rods much more firmly, but the frame does shake end to end unless you use slow acceleration. Surprisingly is seems to have little affect on the printing, at least at the speeds I use.

>That's what I had in mind originally. If it does cause
> problems, I have a plan; I have two 650mm lengths
> of threaded rod left, which I could install at the
> top of the frame (where the Prusa motor mounts
> would normally attach to). I can then triangulate
> onto the frame sides without getting in the way of
> anything. It's certainly going to be wide, but
> should be very rigid - I could even mount the
> filament spool on it. If it all doesn't work, I
> can always revert to motors on top!

I am not sure I follow. To make it stiff I think you need to make the full cube with all the diagonals like Darwin. The threaded rods are only stiff in extension / compression, so you only get rigidity from the diagonals.

I don't think putting the motors at the top helps much, just moving that them to the bottom doesn't gain anything stability wise.

>
> Like you say, the only way to fully constrain the
> axes is in a box.

For the Mendel axis layout you don't need a full box, just two sheets at right angles and two smaller sheets to enforce the right angle, held together with small plastic brackets that B&Q sell, or printed ones if are a purist. That replaces all the threaded rods, nuts and washers!


>I like the Makerbot/Ultimaker
> style because of this, but I don't have a laser
> cutter or router table.

You don't need either to make a box. You can buy wood cut to size as long as it is simple rectangles. You can make a much more rigid box than the laser cut style that Ultimaker and Makerbot use, much cheaper.

>The other advantage of a
> box type printer, especially for a noted supplier
> of parts such as yourself, is that it is
> effectively its own chamber. I also like the
> Ultimaker X/Y axis set up, that they are not
> stacked on each other, and they come together at
> the print head/hot end; it looks very elegant,
> even though it uses more smooth rods. I don't see
> any reason why this couldn't support a
> direct-drive extruder, rather than use a Bowden
> cable. Also, the motors could be mounted on the
> outside, so wouldn't get as hot as inside the
> chamber. Hmm, another day, another project added
> to the list.
>
> Good point about 8mm rod stiffness, but I think
> I'm going to have to keep that standard on this
> one. I'm using SFC8 hardened ground rail, which is
> pretty tough stuff. I saw the saddle shape on
> jcabrer's blog, and wondered momentarily what
> might be the cause.

Several people have stated that harder rods are no stiffer. It's counter intuitive but the people that said so know their stuff mechanically.

>
> > If you mount them vertically the bars will twist
> a little giving a Y error as well as Z.
> Is your concern that the extruder drive weight is
> off-centre? I can see that might be a problem.
> Otherwise I don't see why it should be any worse
> than the existing horizontal rails.

Yes currently the extruder weight is reasonably centred between the two bars so should make them sag similar amounts. I suppose it could be arranged vertically to be balanced as well if the gap is widened to allow the motor to go part way through.

---

[www.hydraraptor.blogspot.com]

Ordered 4 x 1334A and 1 x 1684A (will use this as the extruder motor, and didn't want the shaft sticking out the back) at 11.55pm on Sunday, arrived this morning. So, now I have no excuse for not getting on with my Prusa build... except still waiting for a V9 hot end from mendel-parts.

@nophead - have a look at this new Ultimaker variation, it seems similar to what you are considering, with the motors on the outside like I suggested. Or is that you under a pseudonym?! [www.thingiverse.com]
Could be you, judging by the quality of the prints: [www.thingiverse.com]

No I have kept away from Bowden as I could never see how it would have accurate filament control. Those prints do look very good though.

What I am designing at the moment is like a Mendel, but without all of the frame, just two sheets mounted at right angles. Can't get much simpler than that!

Edited 1 time(s). Last edit at 08/23/2011 01:46PM by nophead.

---

[www.hydraraptor.blogspot.com]