Welcome! Log In Create A New Profile

Advanced

Closed loop vs. stepper motors -advice sought

Posted by jappiemike 
Closed loop vs. stepper motors -advice sought
November 07, 2015 10:40PM
Hello gentlemen,
I have been thinking of my next Arduino/Mbed project and a 3d printer/router/SLS/powder printer seems the go. I've been thinking of building a 500mx500mm welded steel (50x25mm sections) bed machine and using linear encoders along the X and Y axis to give me the accuracy that C7 ballscrews apparently won't. When using encoders why do you use quadrature code when you already know the direction the bed is moving having started to drive the motor clockwise or anticlockwise? Would using smallish DC brushed or brushless, I haven't decided on the pros and cons yet, through a zero backlash worm drive to give the higher torque be the best way to go? Because of the size of the axis would a central driving structure with supported linear rails on the outer extremes be as good as 2 drive structures alongside the rails to combat racking? And finally if I do manage to close the loop can I use a timing belt instead of the ball screws, at the moment I'm not intending to cut anything stronger than wood so there shouldn't be too much of a problem with stretching the belts.

Cheers
Mike

Edited 1 time(s). Last edit at 11/07/2015 10:42PM by jappiemike.
VDX
Re: Closed loop vs. stepper motors -advice sought
November 08, 2015 05:28AM
... quadrature encoding is used to make the pulse resolution of the encoder 4-fold compared to the used lpi density.

With two 500lpi-wheels, arranged at 90deg phase-offset, the encoder gives 2000 clicks per turn.

So you'll need 2000lpi-wheels (what needs much more precise manufacturing) to get the same resolution with single encoding only ...


Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]
Re: Closed loop vs. stepper motors -advice sought
November 08, 2015 04:07PM
There is NO firmware (that I know of) out there that uses feedback loop
reading encoders for positioning.

Are you planning on writing custom firmware?

20 inch linear encoder with good accuracy --- big bucks

How fast is interface to encoder?

Rotary encoder on the timing belt?

confused smiley
VDX
Re: Closed loop vs. stepper motors -advice sought
November 08, 2015 04:21PM
... I'm using servo motors with optical encoders on one CNC-mill (and magnetical encoders on another machine) with standard stepper driving firmwares - the servo-drivers needs only the STEP and DIR signals and handle the feedback on their own. The 'stepping' frequency (encoder speeds) with this drivers can be up to 300kHz without loosing positioning accuracy ...


Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]
Re: Closed loop vs. stepper motors -advice sought
November 08, 2015 04:22PM
There are servo controllers that accept step/direction input...


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Closed loop vs. stepper motors -advice sought
November 09, 2015 04:50AM
Hello gents and thanks for the feedback so far. VDX, which software/firmware are you using, I'm a mac man, so I would like to stay away from Windows if I can.

Cheers
Mike
VDX
Re: Closed loop vs. stepper motors -advice sought
November 09, 2015 05:59AM
... some comercial software/firmware combinations on windows and several of the RepRap-softwares, actually Marlin4Due and Pronterface ...


Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]
Re: Closed loop vs. stepper motors -advice sought
November 09, 2015 07:44PM
Jappiemike Stated "using linear encoders along the X and Y axis to give me the accuracy"

Servo motor with encoder just insures motor moves as commanded
NOT
the accuracy of an end effector driven by leadscrew or belt

It appeared to me that the linear encoder would be independent of drive
indicating where the end effector is located
motor driving until encoder reports location is attained.

Most stepper motor drivers (Easydriver) have DIR, Step, enable, MS1, MS2 (Microstep)

is a servo that much better?
Digital servo is given angle command and it goes to that angle
with accuracy of the measurement device in servo.

I'm confused?

confused smiley



Quote
jappiemike
Hello gentlemen,
I have been thinking of my next Arduino/Mbed project and a 3d printer/router/SLS/powder printer seems the go. I've been thinking of building a 500mx500mm welded steel (50x25mm sections) bed machine and using linear encoders along the X and Y axis to give me the accuracy that C7 ballscrews apparently won't. When using encoders why do you use quadrature code when you already know the direction the bed is moving having started to drive the motor clockwise or anticlockwise? Would using smallish DC brushed or brushless, I haven't decided on the pros and cons yet, through a zero backlash worm drive to give the higher torque be the best way to go? Because of the size of the axis would a central driving structure with supported linear rails on the outer extremes be as good as 2 drive structures alongside the rails to combat racking? And finally if I do manage to close the loop can I use a timing belt instead of the ball screws, at the moment I'm not intending to cut anything stronger than wood so there shouldn't be too much of a problem with stretching the belts.

Cheers
Mike
Re: Closed loop vs. stepper motors -advice sought
November 09, 2015 08:26PM
You're thinking of RC servos, not quite the same thing as a servo motor. RC servos move back and forth over a specific range of angles. A servo motor rotates continuously and has a positional encoder attached to the shaft. The controller drives the motor while paying attention to that ecoder.


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Closed loop vs. stepper motors -advice sought
December 11, 2015 02:36AM
Ben Weiss wrote his master's thesis on adding xy closed loop positional feedback to a hobby printer, check it out: [reprap.org]
Re: Closed loop vs. stepper motors -advice sought
December 11, 2015 06:54PM
Ben Weiss article interesting (a thesis)
and uses Stepper motor.

The link to specifics on encoder [objects.reprap.org] is BAD.

since he mentions "alignment of the optical strip. I use the optical strip from the same printer as the encoder, since the encoder is optimized for a certain stripe spacing."

Probably is Linear encoder

The full pdf version of his paper much better



confused smiley

Edited 1 time(s). Last edit at 12/11/2015 07:11PM by cozmicray.
Re: Closed loop vs. stepper motors -advice sought
December 14, 2015 07:23AM
Hello Gents, I'm sorry for not being more active on this questions but I'm working 14 hour days on a mining site and apart from work eat and sleep I've not got a lot of spare time. I've not read the whole of Ben's thesis but I will get to it soon. I initially thought all i needed to do was alter the axis moving function in the Gcode library and the linear distance reading would "bubble" through the code, maybe I'm wrong, I really haven't had time to investigate. Thanks for all the replies so far, I do appreciate the input on my path of knowledge.
Cheers
Mike
Re: Closed loop vs. stepper motors -advice sought
December 16, 2015 09:48AM
I saw a printer at Maker Faire NYC that supposedly used rotary encoders instead of steppers.

Mod-t 3D Printer

it is a beltless design. the printbed is made out of a material that works well with PLA at room temperature without a heated bed. X and Y were rack and pinion controlled, with the racks molded right on to the solid one piece printbed. The printbed can be lifted off the gears and then put back very easily. they sell additional printbeds if you want them for fast swaps.

The Z axis supposedly uses a rotary encoder motor connected to the drive screw.

I thought it was a very innovative design, but not fully "open" obviously. Print quality was good, but not as high resolution as an Ultimaker. I suspect that this means they used lower resolution rotary encoders to control costs....

I wanted to see what the product looked like inside, but they said they could not open one up.... of course.
Re: Closed loop vs. stepper motors -advice sought
December 29, 2015 10:34PM
My wife's quilting machine uses rotary encoders on the carriage rails to keep track of it's position. It flings an 80lb industrial sewing machine around on a 10' x 2' x/y bed, and can get back to the same position within around 0.01" over that distance. Max speed is around 300mm/second.

It was a big puchase and she has it under a maintenance agreement, so I'm not about to tear into it to learn all it's secrets.

"Gammill" - if you want to try your google-fu. The quilt machines are a very proprietary/closed world, however. [gammill.com]

Edited 1 time(s). Last edit at 12/29/2015 10:36PM by robu.
Re: Closed loop vs. stepper motors -advice sought
February 29, 2016 11:58PM
Check out these servos from Teknic called Clearpath
they are "reasonablly" priced lol at around $350 each

not sure how easily you would be able to use this on a Mega board/ Rambo

[www.teknic.com]
smalled size is Nema 23
I wish they made a Nema 17 size for $200


Makerparts.ca
Your Canadian source for V-Slot Extrusion, CNC and 3D Printing Parts and Accessories.
we are proud to be an official Openbuilds distributor.
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 03:05AM
So, I've decided to go down the path of stepper motors at the moment just to get my machine up and running. I'll use C5 ball screws for now and see how it behaves and then make improvements along the way. I'm still keen on the idea of linear encoders and will look at that later on down the track. Thank you for all the replies to my question.
Mike
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 06:59AM
Can anyone tell me what benefit would come from replacing steppers with the complication and expense of servo motors?


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 09:59AM
I think the biggest advantage is greater torque at high RPM. Steppers give the greatest torque near 0 RPM and torque begins to drop as rotational speed increases. Servos maintain a flatter torque curve with increase in RPM. In the end, if you size the stepper correctly and place the appropriate limits on acceleration and velocity you will have the same reliability as a closed loop system. I have two large CNC mills driven by stepper motors that will drive a 1/2" EM through steel at amazing rates to 0.001" tolerance. Stepper motors - we went to the moon on this technology!
Anonymous User
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 10:31AM
Quote
the_digital_dentist
Can anyone tell me what benefit would come from replacing steppers with the complication and expense of servo motors?

Someone who can answer you, wouldn't envision the use of servo motors/encoders for such thing as a 3D FDM printer, wood working machine or similar requiring an accuracy of 1/10th, moderate speed, no massive parts to move !
Servos, encoders have been in use for 40 years now and yet, steppers are still used (some even fitted with encoders !) despite lower and integrated electronic cost. So ?
yes, servos electronic has moved from analog to digital, but the overall cost still remains high.

Funny but for minimal requirements, like to adjust the height of a car headlights, it is a done with a servo motor and not a stepper. Stepper seems to be squeezed between low end and high end smiling smiley

An other thing to consider, an MDF frame with shining Heidenhain encoders and ball lead screws, the warmth of wood, the glitter of gold and steel, is a thing a beauty isn't ? smiling smiley

Edited 1 time(s). Last edit at 03/01/2016 10:54AM by MKSA.
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 10:51AM
I think the main advantage of using servo-controlled motors to control the XYZ axes of a 3D printer instead of steppers would be that it would avoid the possibility of missed steps.



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].
Anonymous User
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 11:00AM
Quote
dc42
I think the main advantage of using servo-controlled motors to control the XYZ axes of a 3D printer instead of steppers would be that it would avoid the possibility of missed steps.

When I see how most machines are poorly assembled, lack rigidity, if not poorly designed, I bet most of the "missed steps" are due to bindings, parts seizing etc..., in addition of course to asking the stepper too much in term of speed, acceleration ...

With servos, the machine would probably explode smiling smiley

Edited 2 time(s). Last edit at 03/01/2016 11:02AM by MKSA.
Re: Closed loop vs. stepper motors -advice sought
March 01, 2016 11:31AM
Do you still need endstops when using a servo?
or does it just know its limits.

If power is off does remember its position when powered back on?


Makerparts.ca
Your Canadian source for V-Slot Extrusion, CNC and 3D Printing Parts and Accessories.
we are proud to be an official Openbuilds distributor.
Re: Closed loop vs. stepper motors -advice sought
March 03, 2016 01:50AM
Quote
dc42
I think the main advantage of using servo-controlled motors to control the XYZ axes of a 3D printer instead of steppers would be that it would avoid the possibility of missed steps.

Do stepper motors in 3D printers miss steps? I have never seen such behavior in my printer. It seems to me that if a 3D printer's motors are missing steps, they are underspeced for the job they're being asked to do (or there is some mechanical problem that needs to be fixed). You can underspec a servo motor and create problems, too.


Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Closed loop vs. stepper motors -advice sought
March 04, 2016 05:01AM
The occasions I have seen steps lost are usually when an overhanging edge has curled up and the print head has caught on it. This either causes the print to break free from the bed, or steps to be lost resulting in layer misalignment. With a servo-controlled motor, the layer registration would be corrected as soon as the head moved away from the protrusion. The error signal form the servo could be used by the firmware to suspend extrusion while the head is out of position, or to pause the print and alert the operator.



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].
Sorry, only registered users may post in this forum.

Click here to login