Stepper motor alternative ?

I dont have any stepper motors, but i have a lot of something similar, BLDC motors. BrushLess DC motors are cheap, light, powerful, and i have some.

An example of a motor, 130W power, 80g heavy, 12$:
[www.hobbyking.com]
The electronics for it, 20A, 15V 7$:
[www.hobbyking.com]

Now there are some problems. Those motors rev more than 10'000 rpm and dont deliver much torque. They are made for turning and not able to go to an exact position.

The solution for the torque / rpm is a gearbox, gears are plenty and cheap, example here:
[www.hobbyking.com]

The solution for the positionning is more complicated. A BLDC motor is quite similar to a stepper motor. But is has 3 wires instead of 4, it has a different number of coils than a stepper, and the coils are wired up differently. BLDC have a multiple of 3 coils, and are wired a star. The motor above is a 14 magnets, 12 coils motor. This motor, driven trough the 3 wires has 42 steps per revolution. That is 7 repetitions of 6 steps. The ESC (driving electronic above) has 3 half H bridges made with mosfets, is good up to 20 amps current, and has a atmega8 controller inside.

Now BLDC motors do need a position feedback to know when to commutate the coils. This is made with back-emf and does not work at very slow rpm's. I have to run the BLDC motors like steppers and cant use the position feedback. This works, but not with the original firmware in the ESC's.

What i have done is to hook the h-bridges up to an arduino to test the motor. I can control its position, rotate it CW or CCW wise, speed up or slow down, just like a stepper. Once i have it all figured out, and its complicated, i could burn new firmware on the controller and use it as stepper.

Does this have enough torque to drive an extruder? Does it have enough power to drive the XYZ position system? What u guys think ?

Video of a ramp-up speed test.
[www.youtube.com]
1 rotation and direction change:
[www.youtube.com]

I have to increase the current at high rpm or it will stall, and it tends to make a ringing noise going stepwise. Any hints of "stepper control essentials" i could read?

Edited 1 time(s). Last edit at 03/27/2013 03:24AM by Cesco.

Your step size looks quite large and I do not think you will have the required positioning resolution.

Your minimum speed looks quite high, I think this will cause problems as well.

If you search Youtube for videos of RepRaps printing you may be able to get an idea of the speeds and position resolutions we deal with on the extruders.

Honestly, stepper motors are pretty cheap. If you want to make a working RepRap just buy stepper motors and be done with it.

EDIT:

You may be able to get these motors to work by gearing them down very aggressively, with a planetary gearbox of some sort.

Edited 1 time(s). Last edit at 03/27/2013 09:07AM by crispy1.

Minimum step size is 1 step per eternity. Cant go lower than that
Obviously, the easy way is to buy steppers and electonics. I know that. thanks.

But i like this thing:
[www.youtube.com]

Have you seen the work the brushless gimbal crowd has been doing lately?

I'm extra lazy today so I'll just copy/paste my post from another forum.
-------

I think the recent brushless motor based camera gimbal developments in the quadcopter community are going to be super useful for the robotics community in the near term.

[www.rcgroups.com]

In the past few months quite a few low cost, large diameter, and many pole brushless motors have been released for the quadcopter community. The gimbal crowd have taken these motors and rewinded them to work better for closed loop positioning control. I expect in the next few months we'll see a ton of lost cost brushless motors meant for position control along with low cost controllers for them. Swapping out their IMU feedback setup for a magnetic encoder and you have a very good start towards a nice robot servo.
-------

Since I made that post I ran across this guy..
[www.rcgroups.com]
Reading his post 3 he claims that he's not even using a PID and is setting the position by just controlling the field!

Very intresting, thank you for the links. I was not aware of this.

The holy grail would be a position feedback. There are such motors, magnetic position pickup inside the motor, but expensive and wrong KV number.
[www.hobbyking.com]
I need low KV (rpm per volt) motors or i need a huge gearing which will bring lots of slop.

Thanks, im reading

Edited 1 time(s). Last edit at 03/27/2013 02:07PM by Cesco.

billyzelsnack Wrote:
-------------------------------------------------------
> Have you seen the work the brushless gimbal crowd
> has been doing lately?

It looks he has created a custom stepper motor (closed loop) and a custom stepper motor driver. I can't really see that he couldn't have done the same thing with off the shelf stepper motors and Allegro stepper drivers.

I can see that there may be an advantage in terms of weight and reusing components that are already familiar to the 'copter community, but I can't see how that is much advantage to general machine control community.

Cesco Wrote:
-------------------------------------------------------
> Does this have enough torque to drive an extruder?

Best way to find out is to measure it.

> Does it have enough power to drive the XYZ
> position system? What u guys think ?

This question seems to come up every few months. People suggest that theoretically a cheap DC motor can be made into a servo. When you add a gearbox, an encoder and a control unit plus firmware and solve all the problems involved then you will have added a lot of complexity, I am just not sure that the solution will end up cheaper or better. However, you are the first person to demonstrate something working, so that is a positive sign

At the basic level, a stepper motor is a hunk of metal with windings which requires a simple control system, a BLDC motor is a hunk of metal with windings that requires a more complicated control system. Costs to an end user are really about volume of supply, and there seems to be a good supply of cheapish stepper motors. The balance changes when you have larger motors, a $100 motor makes sense to use a $30 controller.

The Achilles heel with DC motors always seems to be getting accurate position/feedback in a cheap form, if you can solve that you may be on to something.

On some old large printers from the 1980's which used server motors they used a PCB instead of an optical disk for feedback. It was simply a circular PCB mounted on the motor shaft with a whole bunch of fine tracks running from the center to the edge. There was a brush on the center and one on the edge. This might be a cheap way to mass produce position encoders??

IMO, there's a time for re-purposing and a time for using. Just buy some cheap steppers and get on with it.

Time is money too... Can't speak for you, but mine costs between $75 and $100 per hour.

As for positional feedback, there are also inexpensive (relative to optical encoders) resistive encoders available. Basically, they're just linear pots with no stop, but for most you'd need dual output motors to drive the encoder.

If time was money then nobody would build a reprap. Time is what you have after you stop exchanging it for money.

bunk. I suppose you pour your own engine blocks too, and cast your own tires. Some of this makes sense, some of it is simply foolishness.

I knee-jerk to the time is money thing. For a very long time I never understood why my dad spent so much time screwing around with our cars. He made more than mechanics so why did he just not take them to mechanics? What I did not factor in is that he enjoys working on cars. $1 at his job did not exchange for $1 working on cars.

Pour my own engine block? Sounds like a pretty sweet project.

The effective cost is the opportunity cost, i.e. what could you do if you weren't doing X. i.e if you spend time trying to re-invent a commodity item like a stepper motor, you could have been innovating something else.

Of course, we all do what grabs our interest, and it is good that that people have diverse interests. If someone does come up with a cheaper alternative to a stepper motor that would be awesome and hopefully we could all benefit.

And I knee jerk to unnecessary moronicisms.

cheaper than a $4 stepper that's a Chinese clone of an American stepper that used to cost $100? lol, you guys are really something.

Please show me where I can buy a $4 stepper.

The same forces that make it possible to push a $100 stepper to a "$4" stepper can make it possible for low cost servo availability.

If you want to make a printer so you can print stuff, you should definitely, definitely just buy some cheap steppers and get on with it (as xiando recommends).

If you're interested in advancing the self-replicating aspect, or you like DIY projects for the fun of it, then it's perfectly reasonable to work on stepper alternatives. There's a whole wiki page on motor alternatives here. Feel free to add new options to that page!

Some of you might like this scratch-built optically-commutated printed-bearing flat-coil BLDC motor. In theory you could make the controller, encoder, and motor coils all part of the same PCB. In some old VCR's and disk drives BLDC motors are made by soldering discrete coils direct to the board. This might be a cheap way to make your own motors, especially if your machine can make PCBs.

The old RepRap DC motor controller board might be a useful starting point for folks wanting to make servos from cheap DC motors. (Although as bobc points out, cheap and accurate position sensing is more of a challenge than finding a cheap controller.)

Oh, and just for those that are interested:
Backyard Casting: water cooled cylinder head w. 4 cores
Manufacturing Tires: modern and in the 1920's

Just saw this one, via hackaday: Homemade 3d Printed Stepper Motor. Not too different from the other printable motors out there but still cool! See also:
printed "4 cylinder" solenoid motor
pcb coil axial air-gap motor
plastic and wood's metal motor

haha that lego parts solenoid thingy !!! so coool !!!

to beat any of these, someone will have to do it steampunked style ! with real steam and a choo choo !

MattMoses Wrote:
-------------------------------------------------------
> Just saw this one, via hackaday: Homemade 3d
> Printed Stepper Motor. Not too different from the
> other printable motors out there but still cool!
> See also:
> printed "4 cylinder" solenoid motor
> pcb coil axial air-gap motor
> plastic and wood's metal motor

---

______________________________________
__my mixed bag blog || aka --> [http] || ___ so 3D printing is everywhere ... dont worry, hospitals can now 3Dprint body parts, they will charge you $1million excluding surgical fees ... you will die paying your debts. thats their aim ___ if every patent expires tomorrow, everybody will surely get a 3dprinter and make EVERYTHING ! ____ there is a "DIY-DTG" t shirt printing forum, you can mod an EPSON printer to PRINT like a pro. ___ CNCzone? overly commercialized it seems ___ my country? they will be taxing you for every cm of road you use and track you to your grave using GPS and its government authorized, now they will fire all the traffic wardens instead.___ EEVBLOG? there is only 1 way to do things --> take it apart like a pro

Quote
redreprap
someone will have to do it steampunked style

It's been done (well, maybe except for "real steam"):
3D Printed Oscillating Steam Engine
Wobbler Engine printed on a 3D Printer

But what's that? You say you want a steam-powered stepper motor? How about one of these with four cylinders:
Radial or Hula Motor

Now just add a fluidic controller and a tape reader and you're all set!
Fluidic Full Adder
Printable Fluidic Logic Element
Fluidic logic 4-bit integer square root circuit based on siphons
3-bit Mechanical Punch Card Reader

At that rpm you could do a huge gear reduction gain a good amount of torque while the motor itself is spining fast enough to use the positional feedback

Well if you want a steampunk stepper motor, I've designed a simple one based around a pneumatic wobble motor:
[cubehero.com]

It uses three inflatable silicone chambers to move a cycloidal ring gear around an inner rotor gear in the same way as the motor described here: [www.act.sys.okayama-u.ac.jp]

It should have about 108 steps per revolution and about 0.9 Nm of torque. It is designed as a prototype, but if it was made bigger and with more teeth on the rotor/ring, it would have the same number of steps as the stepper motors already used by reprap. It will probably be slower though.

I haven't printed it yet, but if the rotor gear and ring gears turn out well, then it should work.

If one could make a fluidic stepper driver for it and a paper tape reader to control it, one would have a RepRap very close to completely replicating all its parts.

I know a place local to me in southern California where I can get stepper motors very very cheaply. I would not be surprised to see them for $4.00 or $5.00. If you're intersted let me know, I can visit the store and tell you what they have.

what about using a wheel encoders for motor control and accuracy?

Just to chuck this out there:
[www.ebay.com]

$6.25, 12V, 300RPM with a built-in gearbox. Not bad, eh? It'll need an encoder, but I think you could find a low-resolution one and attach it to the back of the motor (so the gearbox increases its resolution).

The drivers could also be extremely simple. An Arduino Mega would probably be able to handle the monitoring of whatever encoders we use, and the drivers could probably be two transistors and a capacitor. They might even be able to drop into the Pololu slots on RAMPS.

Just a thought.

Cool. We were kicking around ideas about this in this thread. I would like to make a cheap little "drop in" board that has a step/direction interface, onboard microcontroller (super cheapo), onboard driver, and is pinout compatible with a Pololu driver.

Do you really think the Arduino could handle interrupt driven encoder interfaces for all 3 or 4 motors? I have looked for examples of interrupt-driven encoder interfaces for the Arduinio, but have never found anything that looks like it is the real deal. Maybe I am not looking hard enough... Anyone have any links?

I'm not too knowledgeable about the capabilities of MCUs, but I get the feeling an Arduino Mega could handle four optical encoders -- interrupt-driven or otherwise. If not, we could probably use tow or four smaller MCUs to do that job.

It seems like there are a few new developments since I last looked. This page: Reading Rotary Encoders has two new links - here and here.

The interrupt pins listed for the Arduino Mega are 2, 3, 18, 19, 20, 21. These pins happen to be marked with a dot on the RAMPS 1.4 schematic here.

Ideally both channels of the encoder would go to an interrupt-on-change pin, but apparently you can get OK performance if only one channel goes to an interrupt.

So it looks like the Mega can indeed handle four encoders. It would require pretty substantial changes to the firmware, however.

There is a nice demo in this blog article that shows how to use an Arduino for interrupt-driven closed-loop motor control using the linear encoder and DC motor carriage from an old inkjet printer.

Can a printers optical encoder strip be integrated with a stepper motor?
I'm thinking of how this could be incorporated into the Z-axis of a DLP-SLA printer that could have a lot of backlash due to the use of a non traditional drive mechanism.

RepStrap: X-Axis DC Motor and Linear Optical Encoder repurposed from an old InkJet Printer
[www.youtube.com]