Hello!

Sorry for newbie 1st question, but i need quick & str8 answer.
Building core xy 3d printer in solidworks.
In "Mass properties" my "printhead" + 2 smooth rods are ~5500 grams.

Can 2x nema17 (8401) handle this weight in core xy system?

Thanks & happy new year!

Edited 2 time(s). Last edit at 01/02/2016 10:45AM by zsrevda.

Over five kilograms? Chances are they can move it, but you will have to use a low acceleration, making the setup very slow over all.
The golden rule of printer building is to have as little moving mass as possible

---

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

Hi, first tnx for quick replay.

I didn`t put anything fancy, 2x direct extruder(mk8) with 2x nema17 (4401 ~2x 280gr), 4x sc12uu, 2x e3d`s,
but the fact is that 2 rods + rod ends ( 2x 2 sc12uu+ plates + rod holders), are about 2300gr!!! weight.
I can only spare 1/2 - 2/3 of ~ 1000gr of head plates weight, by choosing different materials (alu, or plastic).
So that is ~500gr.
???

5.5kg sounds way too heavy for Nema 17 motors to me. How large is your printer? I guess it must be quite large, if you need to use 12mm rods instead of 8mm.

---

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].

Print area is ~40x40x40cm... It`s on PC ,so my printer can be that big .... for now

Quote
zsrevda
Print area is ~40x40x40cm... It`s on PC ,so my printer can be that big .... for now

OK so I guess you do need 12mm rods. Let's do the maths:

Holding torque = 54Ncm. But that's at full current and the motors will run very hot. So best accept 85% current and 0.85 * 54 = 46Ncm.

Let's assume you use GT2 belts with 16-tooth pulleys. So the pulley radius is (2 * 16)/(2 * pi) = 5mm = 0.5cm

Belt holding force = 46/0.5 = 92N

This is what the motor will hold, not what it will pull. When accelerating, you probably won't want the motor to be more than 1/8 step behind where is should be. The torque for 1/8 step is 19.51% of the torque for a full step. So the force available is 0.1951 * 92 = 18.4N.

For pure X or pure Y acceleration in a CoreXY system, both motors provide force, so you have double the force i.e. 38.8N. Applying F=ma we get a = 18.8/5.5 = 6.7 m/s^2 or 6700mm/s^2. This is adequate.

Diagonal motion is more critical. For a 45 degree move, one motor provides the force to move both the X and the Y axis. So the diagonal acceleration is one quarter of 6700 in each of X and Y, which means (6700/4) * sqrt(2) = 2370mm/s^2. This is still reasonable.

If I have the maths right, it suggests that those Nema 17 motors should be adequate. Nevertheless, you might want to consider a Nema 23 stepper instead such as [www.omc-stepperonline.com]. This is rated at 1.5A so within the capability of standard drivers, but you will need a 24V or higher power supply (e.g. 24V LED PSU turned up to 28V) to drive it at reasonable speeds, and compatible electronics (e.g. Duet or Smoothieboard). Alternatively, use external drivers, then using 24V power and higher-current motors you can be more certain of reaching high speeds.

[EDIT: corrected the calculation]

Edited 3 time(s). Last edit at 01/01/2016 07:22PM by dc42.

---

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].

Tnx for the math, i`ll consider 23`s, but as you said , that lead me to more$ electronic, drivers...etc. I am aware of that path.
But just tell me simple, can You calculate for me , what is the maximum weight (of parts included in my question, 2rods and printhead), nema 17 can manage @ decent speed
in core xy system scenario?
... so i could be able to stick with ramps+ mega+ 12V...etc.

Most core xy printers use bowden for this reason

---

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

I consider bowden as solution, but for now i would like to stick to direct system, because of possible use of flex filament from time to time.
My current head design have easy switch from direct to bowden already. For now i just need an info from experienced members from this
community on:
How much weight nema 17 (hs8401) can handle @ decent speed in core xy setup?

Out me some weight limits from yours real world experience and/or do some math for me, please?

EDIT: @dc42, Just saw your edit, thanks again. I`ll leave some place (holes) in my top printer plate for nema 23 upgrade, just in case.
For now i`ll stay with 17 setup as you said.

Edited 1 time(s). Last edit at 01/02/2016 06:00AM by zsrevda.

Quote
zsrevda
I consider bowden as solution, but for now i would like to stick to direct system, because of possible use of flex filament from time to time.

Printing flexible filament with a bowden setup is just a question of the right extruder
[www.youtube.com]

---

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

I agree with You, but with direct (or wadeish like) there is no question , just kidding , You are right. I heard other ppl confessions that it can be done with proper extruder /setup, but this is a little offtopic of my Q. Please help me with max weight?

Hi,
dc42 explained very well (thanks again dc42) that my setup will run with nema17 at lower resonable speed,
but i still don`t get it what is the maximum weight of these 2 rods+head, so i can run my future printer at "good" speed.

Sory for my borring question, but i need that info from all of you who knows that kind of calculations better then me.
Here is a picture of my printhead in development.What you think?

I was trying to figure out what you were doing to make the X axis weigh 5.5 kg, and now I see.

You shouldn't need 4 bearing blocks- 3 should fully constrain the motion and reduce the moving mass. Direct drive extruders don't normally need cooling fans blowing on the motors. If the extruders are facing each other that way you'll have a hard time seeing how the filament is feeding when you try to insert filament. You'll also have trouble seeing what's going on if there's a problem.

Have you considered using a single linear guide instead of two guide rails? You will get better performance from the bearings and lower the moving mass at the same time.

You've come up with a mechanical design, but it isn't finished yet. You haven't included any means of connecting the wiring and routing filament to the extruders, hot ends, and fans. I think that once you start looking at that you may end up making a lot of changes to your design. If you're going to use a drag chain for the wiring you'll need some place to anchor it.

---

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

If you want to avoid using a Bowden tube then I suggest you use geared extruders instead of direct drive. You will be able to use smaller, lighter motors and you won't need to run them so close to their rated current, hence no cooling required.

---

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].

Use 4 bushings instead of ball bearings. Scrap the motor fans and use printed extruders instead of metal. For the base plate use carbonfibre or dibond instead of metal.
That should get the weight of the carriage down to way below 1kg.
Due to the lower carriage weight you should then be able to use 10mm rods instead of 12 for the whole assembly.

---

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

Hi,
First of all thank you all. Any sugestions are welcome here, at this stage of developing. All yours deduction are very true also. But my actual design was based uppon
things and parts i allready got from before, (eg. big barrel silicone heater 220vac, a few steppers and bearings) and services that i can get relativly cheap (metal cut).
So my Bom is predestined by WIG.My first printer wolud be all metal, for now. I also purchased from Ali for ~$70 (rapms, mega, drvs,2e3d,1ssr...ect.)
I`ll consider aluminium for head plates.(i coud not get cheap carbon fiber plates, or any carbon fiber plates actually :-), here )
So just tell me simple for now. Would be an kilo,2 or 3kg, the max weight for optimal speed?
Regards.

Quote
zsrevda
So just tell me simple for now. Would be an kilo,2 or 3kg, the max weight for optimal speed?

You simply will not be able to print at what i would care to term optimal speed.
Even fully optimized i doubt your setup will give the same performance as let's say an i3 or an ultimaker.
Even a bit less than 1 Kg is still to much weight to be thrown around with any decent acceleration and you will find that your stepper driversand motors will already work at their limit.

---

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

That`s what i like to hear, thank you for your opinion.
So you said that 1kg would be much enough and the limit for optimal speed/ nema17. What you considered an optimal speed?

But, i done some weight measures. It apears that only 2x 10mm rods at ~550mm lenth + only 3 lm10uu + just 2x E3d without any extruders, are over 1kg.
That parts could not be "in the air". Some plates must be at the ends of the rods , and 4 berings or 2 bushings also + some construction plastic or alu must hold the head?
What about dc42 calculations?
I`m little confused.

Quote
zsrevda
I`m little confused.

I think that is because you are trying to build a printer in a way that is just not geared towards beeing fast.
Your prerequisits define the limits, either go with that and accept that it will not be a speedy printer, or change the concept.
All 3D printers i know that move the complete x assembly along the y axis don't try to put dual extruders on it. Most just use bowden and even then mostly only one hotend.

---

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

Just a note, that is a beautiful drawing and rendering. Nice job.

Thank You mr. elwood127, i am not that bad in drawings, but in physic calculations i am :-).
I still wait for dc42 (and anybody else willing to help here), to resolve my doubts.
I managed to reduce the weight to ~ 3,5kg by slight different design and by using lighter materials for my plates.
I hope that they`ll say that ~3,5kg would be more optimal weight , to be driven at good speed, by that tiny nema17 motors.
But that is only my wishes. I`ll wait to hear some experts opinion.

Although my calculations (if they are correct) suggest that your original design was viable with Nema 17 motors, I agree with everyone else here that your print head has far too much mass. I suggest you start by reducing the mass of the extruder drives by using geared extruders and perhaps Nema 14 motors instead of Nema 17. That in turn may allow you to use thinner rods etc. I haven't built a CoreXY printer so I will leave it to others to provide more detailed advice.

---

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].

There are several areas for concern here, as others have already pointed out 5.5 Kg is a large mass to be throwing around at speed, my Mendel 90 X carriage weighs approximately 800 g, it can print at over 100 mm/s but I restrict it to under 60 mm/s because print quality starts to suffer at higher speeds. Your printer will probably have stiffer construction, nevertheless I seriously doubt it will manage rapid changes of direction at anything like 60 mm/s. Fast printing can only be done with a small moving mass (or with a hugely strong and powerful printer), on the other hand there are limits to how slow you can print, I recently discovered that PLA jambs in my J-head when I (accidentally) printed at only 20 mm/s (the reason is that the molten filament doesn't move fast enough to stop heat creeping up the heatbreak), however maybe something like an E3D head (with active cooling) wouldn't suffer this restriction.

Of even more concern to me for a printer of this size would be the use of unsupported (except at the ends) steel round bars. Unsupported round bars are acceptable on small printers (if you don't want the level of precision that The DD advocates), but they scale horribly for bigger build volumes or heavier loads. To put it simply, if you suspend a weight in the middle of a steel bar supported only at each end, the bar will bend. There are online calculators that let you calculate theoretical deflection, some will even model a realistic spread of weight over the bars, however if we simplify it and assume the worst case (a point load in the middle of the bar), then a load of 5.5 Kg suspended in the middle of two 12 mm steel bars, with an unsupported length of 500 mm (seems a reasonable conservative assumption for a 400 mm print volume) would produce a deflection of 0.36 mm. That is just one axis, in a Core X Y printer you need to sum the deflections of X and Y axis. Of course in the real world the load is not a point load, and the X carriage can actually brace the rails a little, but on my printers I've measured deflection at about 80% of the theoretical figures.

Do you really want a printer that needs a fruit bowl for its "flat" print bed?

I produced the attached spreadsheet a while back to make these calculations, it is what convinced me to abandon steel rods for any horizontal axis on my next build.

Tnx dc42 &Tnx Y all..., wow plankton, very detailed explanation..This become a great topic indeed. More design aproaches is very welcome :-) !
I reduced tottal weight (head+2rods) to ~3,5kg already, and figuring how to lose more with all your sugestions.
@plankton- i think the bending will be less problematic at this new weight also.
More pictures to come...
Regards.

P.S. Future thoughts: z will be light "as feather" with counterweights like in building elevators.( I moved this "thought" to new topic here )

Edited 1 time(s). Last edit at 01/10/2016 12:39PM by zsrevda.

Quote
zsrevda
P.S. Future toughts: z will be light "as feather" with counterweights like in building elevators.

You are aware that this will not help if you are not working against gravity? There is no free lunch in physics

---

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

Quote
i think the bending will be less problematic at this new weight also.

Less yes, but still far too much if you actually plan to use all of that 400x400mm print bed area - how can you print at 0.2 or 0.3 mm layer height, when your print head height varies that much over the width of the bed?

Quote
z will be light "as feather" with counterweights like in building elevators

excellent, that's the weight dealt with... now what to do with all that pesky mass

Quote
Srek
You are aware that this will not help if you are not working against gravity? There is no free lunch in physics

Explain this a little bit here please.
Also, I really don`t understand this "working against gravity" and what is the correlation with "free lunch in physic"?
English is obviously not my native language :-).

I moved this "future thoughts" to new topic here.
This is just because, i would like to stick here, to xy weight properties and possible design solutions.
Please understand that, as i really happy to learn from you and others members ofcourse!

Quote
plankton
excellent, that's the weight dealt with... now what to do with all that pesky mass

Edited 2 time(s). Last edit at 01/10/2016 12:41PM by zsrevda.