I'd like to build a 3D printer, and I'd like to start with choosing the best Controller...

So I'm about to start a 3D printer project,
And after much time spent online I get the sense that your Controller / Firmware / Stepper Drivers / Stepper Motors all have to "play nice".
And if you purchase all of these components blindly, chances are they wont work well together.

That being the case, it seems to make perfect sense to start at the "brain" (Controller) of a 3D printer and go from there.
I'd appreciate your feedback in helping me choose the right controller for my 3D printer project.

Here is what I'm looking to build, (as related to choosing a controller):

12" x 12" x 12" build-volume
heated bed.
dual extruders.
dual hot-ends.
2 x-axis steppers
2 y-axis steppers
4 z-axis steppers (because I'd like to make a printer that can be easily scaled-up to say a 24" x 24" x 24" build-volume)
SD card printing
LCD control (not a critical requirement for me, but it would be nice to have that option available)
6 end-stops (2 for each axis) *3 may work, but 6 seems safest.
open-sourced with an Arduino backbone would be nice.
(although I've read a bit about 16 Mhz CPU speed starting to become a limiting-factor, and perhaps that would be a problem for me if I wanted wicked-fast printing?)

On a "related" note:

The most similar printer to what I would like to build seems to be this one from QU-BD:
[store.qu-bd.com]

And they are using an Azteeg X1 (not sure why they don't use the X3)
So I started looking at the Azteeg X3:
[www.panucatt.com]
[reprap.org]
Which can apparently control up to 8 steppers.
- 5 right off the bat,
- and a shield to control 3 more.
But strangely enough, if you take a look at the wiring schematic:
[files.panucatt.com]

It suggests paralleling/serialing 2 z-axis motors off of a single z-axis stepper-controller, which I find a little strange considering the controller claims control for "up to 8 steppers" (with the shield).

I found this very helpful:
[reprap.org]

but it seems to lack "# of steppers" as a comparison factor.

I'd love to know what you guys think about controllers,
and which controller you think would be best for me (and why).

Thanks guys!

Almost all the Reprap electronics drive both of the Z steppers in parallel.

Part of the reason is that due to the way they are driven, if one fails to step, the other ALSO fails to step, so your Z usually stays level. If you drive them independently, and one fails to step a few times, your system will be unbalanced and could lead to binding.

Q: What are you trying to achieve by using twice as many steppers? The issue of build volume is usually speed. The reason it takes longer to print a larger object (ie: scaled up) is that the amount of travel distance you need to perform with the head is cubed, and it's very hard to push plastic out of a nozzle very fast in a quality way, so you tend to have to slow your printer down. You usually don't need that much power to require more than 1 stepper to drive an axis unless you have a VERY HEAVY machine. The only reason most designs use 2x Z motors is because a second motor is about the same cost as using a bunch of pulleys, bearings and a belt to connect both (or more) Z rods. It also avoids the need to allow for the need for space for the belt between the 2x Z's.

By using more than one extruder, you can however at least double the amount of plastic you push out. There are designs that use dual extruders on the same X axis (independent movement) that could print two halves of the same object at once (if you can avoid collisions around the middle, which the slicer 'may' be able to do for you).

If you were to try and do dual (or more) extruders using both X and Y using bed that moves in Z design (like the QU-BD that you linked), apart from the issue of avoiding any potential collisions (which would grow by a square law), you would need a way to try and make the extruders avoid all the mechanics for the other extruder (eg: any gantries).

Thanks for the response,

I must not have been very clear with my wording, because I don't plan on having any of the hotends moving independently. The "dual x" and "dual y" motors would be for increased speed and better symmetry (pulling from both sides of the gantry, similar to the reason for repraps to push the z-axis from two sides). As far as I know, the QU-BD printer is single extruder, single hot-end... With no chance of collision.

I suppose on very large prints, a dual hot end (or 3, or 4) could lay down twice (or 3x, or 4x) the plastic at its designated offset distance. But that's not what I'm concerned about right now. I'm mostly interested in speed and accuracy.

Hope that helped clarify.

As mentioned above, your speed is limited by the plastic rather than by the speed / power of the motors. Positioning accuracy will go up with things like fancy drive screws (except for their coarse pitch). Print accuracy (like speed) will be limited by the plastic coming out of the hot end. For a more accurate printer you would go to a 0.2 mm hot end vs a 0.4. With a small hot end like that, your full volume prints are going to take a *long* time. All screw drive is also not going to be very fast ....

If you run multiple motors simply for symmetry, the parallel approach works for them. You only need one driver per axis plus one for each extruder. There are a number of controllers that will do that quite well.

Thanks for the tips.

I plan on running belts.
And I would probably go for .5 mm hot-end for what I do,
but wouldn't mind having one of the nozzles be .35 mm (or .2 mm).

What are the downsides to running multiple motors in parallel off of a single stepper-driver?
It seems to me that the biggest downside would be a potential power (current) limitation.

The other option seems to be to run multiple stepper-drivers in parallel with the controller,
But Cefiar points out that this configuration could prove problematic if steps were to be missed.

I guess the next question is:
What's the likelihood of missing steps?
And when we talk about missing "steps", is that the same thing as missing "micro-steps"?
For example, if a stepper were to miss a "micro-step", does it correct itself at the next full-step? (For some reason I'm picturing "full-step" positions as check-points for micro-stepping, which may very well not be the case...)
And if said motors are linked with a belt (as shown in the QU-BD printer, what effect would one of them missing a step have on the other?... hum...)

Thanks again guys for the feedback.
Still not sure which controller I should start with,
(or whether or not the controller is a good place to start)

I don't think driving two steppers will make it run any faster. Screw drives have backlash problems (play in the screw mechanisms) unless you go to adjustable nuts and funny screw threads (ACME etc) then you have servicing problems. Stepper speed is a frequency limiting factor governed by current in the coils (and microstepping won't make it quicker - you're still up against change in time to drop held current). Dual Z- motors seem to be about keeping the gantry balanced (stop it cockling over and locking in place) and the reduced current in the coils means you can actually turn them quicker. You can also increase speed by upping the voltage.

I can run mine at about 400RPM on the X/Y axis and because I'm running half the current through each of the Z-axis motors, they can run roughly twice the speed

[www.daycounter.com]

I use this site to calculate the max speeds and it's worked for my stuff so far (running Nema14's at 250mm/s, slightly under their max speed)

When you are running in micro stepping mode, the stepper it behaving like an AC motor. Missing a micro step is not a problem if it occurs between the motor and the driver (but- see below). If it occurs between the controller and the driver, lost is lost.

To easily loose a step between the motor and driver, you either need to open the wire to the motor or short it out. Either way there’s a lot more than just the micro step being thrown out. You are abruptly taking the motor current to zero. That’s not the same as simply holding the previous value a bit too long. Your motor will *not* be happy when you suddenly do that.

I would not start with the controller at all. Electronics are (relatively) cheap and flexible. I would work out the mechanical side of things first. That will get you to speeds and torque needs. That drives the motor size selection. With any size you can then pick a voltage and current combination *along with* a driver. Then the controller will come into play.

so can any controller work with any stepper-board?

There are hundreds of different controllers and probably thousands of different stepper boards. People have been making both since at least the 1940's. There are plenty of combinations that will not work together.

If you get a modern stepper driver that will accept step forward and step back signals as pulses, then it can be made to work with most controllers. There are plenty of those sort of controllers out there and they are pretty cheap. You want one that a driver that will do micro stepping, so that rules out a lot of the really old stuff. You want a controller that is based on a modern CPU, so that also rules out a bunch of old stuff. I would suggest that you also want an open source based controller. They all seem to do things the same basic way.

So the real answer is that most of the controllers you would actually consider will work with the drivers you would actually consider.

I'd second Uncle_bobs suggestions - work out the mechanics first, then choose motors to drive it all (you'll be surprised what a single motor is capable of), a good mechanical design will not require masses of force/power to drive it all, especially as we're talking about moving a relatively light weight in what should be smooth straight motion (unless you're talking deltas, in which case, your mechanical design is even more important)