1

Topic: Dual "Y" drive motors

I was looking over Wompus's post where he moved the "Y" stepper motor outside the case and put it in line with the drive rod, and this got me thinking...

Has anyone considered using 2 stepper motors to drive the "Y" belts directly?  There are at least a couple of ways of mounting the motors.

One would be from the outside of the frame with the shafts sticking inside (would require motors with longer shafts).  Another way would be to make some kind of mounting bracket (plastic or metal) that would let us mount the motor on the inside back face, with the shafts pointing towards the left and right side.

The big question is whether or not we could drive the 2 motors from the same output port. 

Hopefully some of the electronics wizards here might be able to chime in on this.  Can the stepper drivers handle 2 motors?  And this question pertains to both boards, Sanguinololu and Printrboard.

To print or, 3D print, that is the question...
SD3 printer w/too many mods,  Printrbot Simple Maker Ed.,  FormLabs Form 1+
AnyCubic Photon, Shining 3D EinScan-S & Atlas 3D scanners...
...and too much time on my hands.

2

Re: Dual "Y" drive motors

Printrbots have two z motors running from the one outputs, with a special connector that splits from one plug to two. The stepper driver current is turned up to compensate. That should be easy to source online.

With the steppers facing outwards as you have stated, they would be turning in opposite directions so one would need the wiring reversed.

Apart from that the physical mounting would be the only difficult part.

3

Re: Dual "Y" drive motors

Also, "Y"? (why?)

I don't see the Y-axis being very long to have a problem driving it from one side.

4

Re: Dual "Y" drive motors

Tomek wrote:

Also, "Y"? (why?)

I don't see the Y-axis being very long to have a problem driving it from one side.

It's not too long, the usual reason is so you don't have a shaft coupling the two sides and you get more torque to move the heavier axis.

5

Re: Dual "Y" drive motors

Yes, you can run dual steppers, but, you are limited in that you still only have a 2 amp use able max on the Stepper Drivers unless you have special high current multilayer drivers, or a massive amount of active cooling. So your total torque is often not much improved vs a single decent stepper motor. It's more about the link rod than anything.

Running multiple drivers can introduce sync issues, so also has issues to consider.

Consider also this will increase current needed on the board as well, meaning you are really pushing any under spec/over driven connectors and tracks.And on a RAMPS board will put you over the 5 amp limit on steppers unless you replace the PTC fuse or pay attention to tuning your total current on all axis.

6

Re: Dual "Y" drive motors

Adrian is right about the current issues on the printrboard, my response was more towards the generic use of duals' vs. a single motor as I have a couple of cnc's that use dual y motors but they also have dual drivers.

7

Re: Dual "Y" drive motors

What about using smaller motors?

The Palolu stepper drivers (and the ones on the Printrboard) are rated at 2A.  The average NEMA 17 stepper is rated at 1.5A+, so obviously 2 of those would seriously overheat the stepper, but what about using 2 smaller, NEMA 14 stepper motors?  The average NEMA 14 stepper draws 1A or less, so having 2 of those should not overheat the driver, and having 2 of them should make up for the lower torque when compared to the NEMA 17.

To print or, 3D print, that is the question...
SD3 printer w/too many mods,  Printrbot Simple Maker Ed.,  FormLabs Form 1+
AnyCubic Photon, Shining 3D EinScan-S & Atlas 3D scanners...
...and too much time on my hands.

8

Re: Dual "Y" drive motors

The issue with doing this without a linking mechanism is the motors can get out of step with each other, skewing/jamming the axis...

9 (edited by Tomek 2013-11-18 00:12:36)

Re: Dual "Y" drive motors

I assumed the dual Y-axis would still be linked, as that way you would not really have a step issue. It seems that with two small unlinked motors you risk skewing without any clear benefit.

At a certain size, though, it can make sense. You would want some slightly high voltage rating/lower current steppers, and drive them with an increased voltage to overcome their inductance. But you would still want the two steppers on the same board. I don't know what the voltage limit is though on the control boards we use.

Our motor drivers are current-limited devices, so we can run a stepper that is rated to a higher voltage and get a corresponding increase in driven power. But higher voltage steppers at the same size are going to be higher inductance, which dictates how quickly you can step/run them for a given voltage.

Anyway, just throwing some stuff out there. overall seems everything is covered

10

Re: Dual "Y" drive motors

Lets bring back the original (more pertinent) question: Why bother? Where is the payback? What is so wrong with the Y axis that is being solved by this design change?

What is the gain when you consider the investment: Purchasing another stepper motor, possibly upgrading the stepper driver, making new mounts, wiring this all up, cutting the drive shaft or buying two smaller ones, modifying the frame to install all of this.

So what is getting solved?

Chuck Bittner is a quadriplegic gamer who is petitioning the major console developers to include internal button remapping in all console games. You can help.
Sign Chuck Bittners petition

11

Re: Dual "Y" drive motors

Hazer wrote:

Lets bring back the original (more pertinent) question: Why bother? Where is the payback? What is so wrong with the Y axis that is being solved by this design change?

What is the gain when you consider the investment: Purchasing another stepper motor, possibly upgrading the stepper driver, making new mounts, wiring this all up, cutting the drive shaft or buying two smaller ones, modifying the frame to install all of this.

So what is getting solved?

Why?  Why not.  We make all kind of mods to our printers, and a lot of them are unnecessary anyway.  So why not this.

But since you asked...

The idea behind this is that most of our problems with getting nice perfect circles is in the slight misalignment of the "Y" belts.  If one belt has more/less tension than the other or is not lined up almost perfectly, or the connecting shaft slides left or right, it creates an imbalance which contributes to the 'skewing" or the flat spots we see when we print circles.

Using 2 motors running in sync, that drive the belts directly without any additional interconnecting rods or extra drive belts, should in theory provide a perfectly synchronized, carriage.

As for the complexity of it, it's not as bad as you make it sound.  2 NEMA 14 stepper motors (~$12/ea.), no rods (the belt sprockets are attached directly to the motor shafts), a Y splitter to connect the 2 motors. 

The only "real work" is getting the motors mounted, and even that is pretty simple.  I could get some aluminum angle from the hardware store, cut it drill it, bolt it to the back face of the SD frame, and bolt the motor to it.  Or print my own mounts.

I think that the only thing to truly worry about is the motors getting out of sync.  That would not be a good thing, but the only way that would happen is if one motor is failing for some reason.  Since both are driven by the same driver, they both get their electric pulses simultaneously, so it's not a matter of them getting out of sync because of some external problem (board or stepper driver related), only some internal problem in the motor.

If you look at the wiring schematic for the Sanguinlolu board, they show it with the optional splitter for dual "Z" motors.  I understand that "Z" motors don't work as hard as a "Y" motor, but again, as long as I stay under the 2A limit, why shouldn't it work?

So this is my long version to the question "WHY BOTHER?"

I'm not saying I'm going to do this anytime soon, but it's something worth exploring, if nothing else, get the brain cells engaged.

To print or, 3D print, that is the question...
SD3 printer w/too many mods,  Printrbot Simple Maker Ed.,  FormLabs Form 1+
AnyCubic Photon, Shining 3D EinScan-S & Atlas 3D scanners...
...and too much time on my hands.

12 (edited by adrian 2013-11-18 02:32:24)

Re: Dual "Y" drive motors

If this is about circle alignment, I've never had an issue from the day I added a pillow block support to the link rod to stop flexion.

Even with grossly mis tensioned belts I have perfect circles. And circles parent impacted by left/right tension, it's 99% over tensioned y-motor to link rod. I believe you are fixing the least significant side of the problem.

Just making the point if this is in pursuit of circles the dual motor approach is over thinking the actual problem. 2 bearings and a 0.02c printed part solves the issue permanently.

But don't let any of us stop you on your journey if its tinkering for the sake of tinkering.

13

Re: Dual "Y" drive motors

The twin z steppers aren't physically linked on the printrbot and it doesn't seem to be a problem.

14

Re: Dual "Y" drive motors

lawsy wrote:

The twin z steppers aren't physically linked on the printrbot and it doesn't seem to be a problem.

Screws can't be backdriven, belts can. If you bump a dual belt system with the motors off... welp, see ya later!

15

Re: Dual "Y" drive motors

pirvan wrote:

The idea behind this is that most of our problems with getting nice perfect circles is in the slight misalignment of the "Y" belts.  If one belt has more/less tension than the other or is not lined up almost perfectly, or the connecting shaft slides left or right, it creates an imbalance which contributes to the 'skewing" or the flat spots we see when we print circles.

Using 2 motors running in sync, that drive the belts directly without any additional interconnecting rods or extra drive belts, should in theory provide a perfectly synchronized, carriage.

I asked because I knew the answer. And the design is wrong for that answer.

If this is about circle alignment, I've never had an issue from the day I added a pillow block support to the link rod to stop flexion.

Even with grossly mis tensioned belts I have perfect circles. And circles parent impacted by left/right tension, it's 99% over tensioned y-motor to link rod. I believe you are fixing the least significant side of the problem.

Just making the point if this is in pursuit of circles the dual motor approach is over thinking the actual problem. 2 bearings and a 0.02c printed part solves the issue permanently.

Let me add to that: A solid shaft is more accurate than any possibility of breaking the system apart. There is no backlash or flexion from one end of the shaft to the other end. But that's moot. Your trying to solve the problem by changing something that is not even part of the problem. As adrian pointed out, perfect circles comes down to tension of the 2 Y belts (which your design does not address at all) and that also being tensioned in sync with the X belt (another part not being addressed by 2 Y motors). You can do all this work, and your problem is still completely unchanged.

You want a good solution to perfect circles? COREXY. A single belt driven by 2 motors for both X and Y axis. You have a single equalized tension in both directions, and your tension does not have to be perfect either (just enough to remove slack, but not enough to stretch). Another benefit is that the 2 motors are statically mounted to the frame, which means there is no motor adding weight to the X axis. Which means lighter carriage assembly, less torque, and less motor heat.

Hell, I have been thinking about doing this with all original SD3 parts. Use the Y and E motors for the CORE, and then use the X motor as the extruder with a Wades geared setup. Purchased parts would be 2 GT2 pulleys, 624ZZ bearings for idler pulleys, 625 pulleys for the X ends, GT2 belt (open), and a dozen LM8UU (get twice what you need to pick out the good ones). All of that is less than $20 from robotdigg. The wades would increase the steps/mm and the torque, and help get rid of that moire effect (although at the cost of speed, but you dont really need it on the extruder).

That has the potential of getting payback for the effort.

Chuck Bittner is a quadriplegic gamer who is petitioning the major console developers to include internal button remapping in all console games. You can help.
Sign Chuck Bittners petition

16 (edited by pirvan 2013-11-18 05:50:40)

Re: Dual "Y" drive motors

Well then... I guess I should just give it up then... and I though I had come up with the answer to life's problems. sad

C'mon guys.. where's the spirit of adventure?  And what's wrong with turning our machines into the ultimate Rube Goldberg machines just because we can  smile

Hazer wrote:

You want a good solution to perfect circles? COREXY. A single belt driven by 2 motors for both X and Y axis.

Uhhh...  you might want to get a second belt just in case. tongue

To print or, 3D print, that is the question...
SD3 printer w/too many mods,  Printrbot Simple Maker Ed.,  FormLabs Form 1+
AnyCubic Photon, Shining 3D EinScan-S & Atlas 3D scanners...
...and too much time on my hands.

17

Re: Dual "Y" drive motors

sometimes it's cheaper to get one longer belt and cut it into two suitable lengths... depends on the type of belt wink

18

Re: Dual "Y" drive motors

ronsii wrote:

sometimes it's cheaper to get one longer belt and cut it into two suitable lengths... depends on the type of belt wink

True enough.

To print or, 3D print, that is the question...
SD3 printer w/too many mods,  Printrbot Simple Maker Ed.,  FormLabs Form 1+
AnyCubic Photon, Shining 3D EinScan-S & Atlas 3D scanners...
...and too much time on my hands.

19

Re: Dual "Y" drive motors

Uhhh...  you might want to get a second belt just in case.

My bad. Firmware calls it COREXY. What I am talking about is HBOT. Single belt.

Chuck Bittner is a quadriplegic gamer who is petitioning the major console developers to include internal button remapping in all console games. You can help.
Sign Chuck Bittners petition

20

Re: Dual "Y" drive motors

I think CoreXY refers to the part of the software that does the kinematics for an Hbot. It's really two belts though.

http://corexy.com/theory.html

21

Re: Dual "Y" drive motors

HBOT is a single belt.

Chuck Bittner is a quadriplegic gamer who is petitioning the major console developers to include internal button remapping in all console games. You can help.
Sign Chuck Bittners petition

22 (edited by Tomek 2013-11-18 20:05:17)

Re: Dual "Y" drive motors

Hazer wrote:

HBOT is a single belt.

Did you look at my link?

I think it's a matter of perspective. I tried to look up H-bots, like http://www.youtube.com/watch?v=dVMA883Quok, but that's two belts not one. It's just two belts acting in unison. 

At least, if you're making this claim, please help me understand the difference between an Hbot and the link I sent [the CoreXY one] I am having trouble understanding


EDIT: Below it is clarified [I am wrong]

23

Re: Dual "Y" drive motors

CoreXY and H Bot are slightly different. They look the same, but have slightly different belt paths. Each have their benefit.

CoreXY has two shorter belts, all four ends attach to the x gantry.

H bot only has one, and so only two ends are attached to the gantry.

CoreXY's belts are shorter, so in theory it should have less backlash, but the belt has to cross over (like in the video) for it to work. I personally prefer coreXY over h bot, but that's just my opinion.

24

Re: Dual "Y" drive motors

Thanks for the info! I will have to find more info about H-bots, I don't fully understand them nor have I found a good drawing/diagram.

25 (edited by Hazer 2013-11-18 20:40:41)

Re: Dual "Y" drive motors

pretenda wrote:

CoreXY and H Bot are slightly different. They look the same, but have slightly different belt paths. Each have their benefit.

CoreXY has two shorter belts, all four ends attach to the x gantry.

H bot only has one, and so only two ends are attached to the gantry.

CoreXY's belts are shorter, so in theory it should have less backlash, but the belt has to cross over (like in the video) for it to work. I personally prefer coreXY over h bot, but that's just my opinion.

I will add to that: Since HBOT has a single belt, belt tensioning has less of an impact on the system (perfect circles). The bad thing is the X axis flexing out of square. As the distance between ends increases, the linear rods have less chance of being able to keep the axis square to the Y axis with fast movements. A good compensation for this is using linear rails instead of rods.

A COREXY has the benefit of using the tension of the two separate belts to counter-act the problem with keeping the X axis square, but requires both belts to be perfectly tensioned to each other.

I prefer HBOT on the merit that it is easier to setup. You can use proper design to keep your axis square, and the single belt makes taking things apart repeatedly much easier. You dont have to tighten the belt at the ends either, if you put oblong mounting holes for the motors so you use the motor to tighten the belts (like the Y motor on the SD).

Here is a good illustration of an HBOT:
http://www.thingiverse.com/thing:161000

Chuck Bittner is a quadriplegic gamer who is petitioning the major console developers to include internal button remapping in all console games. You can help.
Sign Chuck Bittners petition