?

I found this interesting piece on driving stepper motors at speed:

Using a cheap $5 92oz Nema 17 stepper, I was able to achieve around 600 rpm using the misfit tech board.? Wanting to improve performance, I changed out the driver using a TI DRV8823 and 24v at 1.5amps per phase max with PSOC 5LP as the driving MCU updating the PID loop at 1Mhz (1us loop time) and the encoder in ABI mode gives me 2500rpm in 90 degree advance and 9,900 rpm (165 rotations per second) at 180 degrees out of phase.? More analysis needs to be done as there are some special edge cases where the motor ends up going backwards when driven at 180 degrees but it's clearly possible.?


Ant?




On Tue, Jan 7, 2020 at 02:10 PM, John Dammeyer wrote:

Micro-steppers tend to run out of torque at around 700 RPM in general.? It's the physics of the design.? Two things have to happen to make them turn.? You have to change the direction of the current through the windings.? The back emf from the RPM which is what limits standard DC motors is the second problem.

?

Think of it this way.? Turn a DC motor at a specific speed and as a generator it produces a voltage.? If the applied voltage is say 12V then that generator voltage rises to 12V and that's what holds the motor speed to a given RPM. ?Increase the applied voltage to increase the speed until the back emf matches again and the speed stays steady.

?

Stepper motors have very low voltage windings and use current feedback to determine when to shut off the voltage.? That's when the drive starts chopping the power to hold the current at a given value and that torque (amp x turns) is what you get to rotate the armature.? But now you have to reverse the current through the windings and if the motor is turning it's generating that back EMF.?

?

Plus the current through the winding has to decay (inductance slows that) before it can reverse. The more pressure? (voltage) the faster the change in current.? That's why stepper motors with 1.5V windings have 24V power supplies.? But at a certain RPM the back EMF matches that power supply voltage and it takes too long to reverse the current in the windings before the next step.? End result no torque. Motor locks up.

?

So even with an encoder, you still have speed limitations.? And, BTW, the motors with encoders come in larger than size 17.? I used one with a gearbox on the end of it as size 23 and 400oz-in with an encoder built in.? Used CANopen messaging for control. Real sweat motor.? Still not more than about 700 RPM before there were issues.

?

OTOH, the Bergerda AC Servos down side is they require 220VAC so you can't run them on 110VAC or a 24VDC battery system.? But with the 2500 line encoder they run 3000 RPM.? I swapped on in place of a 90VDC DC Brushed Servo. ?Worked well for the mill X axis at 150 ipm.? (3:1 to 5 TPI lead screw).? And not nearly as mechanically noisy as steppers.

?

John

?

?

?

?

From: [email protected] [mailto:[email protected]] On Behalf Of Ant No
Sent: January-07-20 1:18 PM
To: [email protected]
Subject: Re: [digitalhobbyist] Step #1 Encoder

?

I'm surprised, what speed is that?

Smooth and quite is what the the new driver modes do well using clever microstep control. Spread, stealth, chop.

???


On Tue, Dec 17, 2019 at 03:07 PM, John Dammeyer wrote:

But then there is the problem with steppers is they just can't turn as fast as you need

?

Ant, do you have a link for " Variable step rates controlled on the fly by SPI are a common feature on drivers currently available."? What drivers? Everything I am seeing are only set by switches when the motor is not turning or powered. I need 3 - 5 amp drivers and would love dynamic? step control.

?

Ralph

On Wed, Jan 29, 2020 at 10:16 AM Ant No <cmpaqp1100@...> wrote:

If I've understood correctly you can beat the 700rpm speed limit by changing from microsteps to full steps? Does it take full steps or could you get away with a low microstep rate such as two or four? Variable step rates controlled on the fly by SPI are a common feature on drivers currently available.

Seems as though gearing for increased tourque is also gearing for increased precision. Obviating the need for microsteps? Obviating the 700rpm limit?

What I find interesting about these new closed loop drivers is that they use a simple diametric disk magnet (North and South are diametrically opposite sides of the disk, not the flat faces of the disk) with a clever sensor chip that measures multiple magnetic field properties to create the equivelant of a 32k line per rev encoder.

Seems like that should be transferable tech. You can buy the driver without the motor now.

The only issue can be getting the diametric disk precisely centred on the end of the motor shaft. Prolly trivial for the skill sets you have.

There is debate how essential this is. The calibration procedure sets things up after all. People claim that when used for 3d printing, the cheaper MKS motor sets, selling for $20, produce a moiré surface artifact known as salmon skin.

Speculation varies from thinking centering the magnet, PID tuning,? or skimping with a different magnetic sensor chip and? less well filtered circuit are the cause.

The $50 ones from originators, MisfitTech don't appear to have the problem.

They use this sensor chip:

The TLE5012B is a 360° angle sensor that detects the orientation of a magnetic field. This is achieved by measuring sine and cosine angle components with monolithic integrated Giant Magneto Resistance (iGMR) elements. These raw signals (sine and cosine) are digitally processed internally to calculate the angle orientation of the magnetic field (magnet).

I'm not sure it matters in this application anyways but...

Further research while writing this provided a possible cheap fix for the cheaper motors. Eight diodes per motor avoid the moiré effect. Was for same effect on different driver though.



There are a couple of alternative explanations of why it works in the comments. I'd be interested in any opinions anyone here has on which is the better explanation.

This method appears to provide? high resolution? encoder feedback at trivial cost. Eliminating the need for expensive optical encoder disks and replacing them with a more robust solution.

Looks like the future of servo feedback to me.

Ant?

?

?

--
Clausing 8520, Craftsman 12x36 Lathe, 4x12 mini lathe, 14" Delta drill press, 40 watt laser, Consew brushless DC motors and a non working 3D printer

?

John,

Is the round wheel spring loaded? In other words, does it have a center that it returns to? Is the center wheel set up like a mechanical hand wheel on the lathe or mill, or does it move the axis faster as you move the round wheel further off center. Just trying to get in my head options on how different devices are going to control things.

Dale

I use a ShuttleExpress on my CNC Router.? Love it.? Much easier to use than a pendant from China which I also have.? A friend has the larger Shuttle that has extra buttons.? Not sure all those would be needed on a lathe.

?

?

John Dammeyer

?

?

From: [email protected] [mailto:[email protected]] On Behalf Of Dale Grice
Sent: February-06-20 7:59 AM
To: [email protected]
Subject: [digitalhobbyist] MPG & Mach4

?

New to CNC. I have a pokeys57 running a mini lathe. Took a while to get working. Now I would like to get away from a PC keyboard and mouse. I have been looking at many different MPGs. It seems that almost all work with Mach3 and not Mach4.

So I throw it out to the group. Any suggestions for a neophyte?

Thanks,

Dale Grice

I use a ShuttleExpress on my CNC Router.? Love it.? Much easier to use than a pendant from China which I also have.? A friend has the larger Shuttle that has extra buttons.? Not sure all those would be needed on a lathe.

?

?

John Dammeyer

?

?

From: [email protected] [mailto:[email protected]] On Behalf Of Dale Grice
Sent: February-06-20 7:59 AM
To: [email protected]
Subject: [digitalhobbyist] MPG & Mach4

?

New to CNC. I have a pokeys57 running a mini lathe. Took a while to get working. Now I would like to get away from a PC keyboard and mouse. I have been looking at many different MPGs. It seems that almost all work with Mach3 and not Mach4.

So I throw it out to the group. Any suggestions for a neophyte?

Thanks,

Dale Grice

I used Clough42 because it is very well documented both in hardware
installation and software design and covered my needs. It is a simpler
system than most (for instance it has limited UI options based on the
8 digit display and some fixed LED) with less functionality as a
result.? I will likely improve that over time, maybe adding a better
display and more functionality.? The good documentation (both written
and through YouTube videos) made it pretty straight forward to figure
out how to adapt it to my lathe.

It's also pretty easy to swap out the actual control logic between
these ELS systems.? The microcontroller and code are the cheap part of
the system, the stepper or servo drive and custom brackets and pulleys
is a bigger investment in money and time.

alex
------------------------------------------------------------------------
*From:* [email protected] <[email protected]> on
behalf of Richard via Groups.Io <edelec@...>
*Sent:* Monday, February 10, 2020 11:16 PM
*To:* [email protected] <[email protected]>
*Subject:* Re: [digitalhobbyist] RELS project info wanted #RELS
I have put the Sketch and a Spreadsheet in the "Files" area. The
Spreadsheet is used to generate the system parameters for "Your" lathe.
This is based on Metric values only for the moment. However the only
time you might be aware of that is that the "Feed" pitches are in
Metric. The unit will happily install and run on an imperial machine.
Metric and Imperial thread pitches LH and RH are both available, Manual
and Automatic, together with the ability to cut any taper Internal or
External.
I will be doing a write up on the system to explain what it can do in
the near future.
I would be interested to know why you picked on the Clough42 system for
your lathe.
Good Luck
Richard


On 10/02/2020 16:02, alexphredorg wrote:

Is there a GitHub or other page for the RELS project being discussed
on this list? ?I think I’ve read all of the posts but never found a
summary page or link to the source code.

I’m currently adding an electronic leadscrew to my Emco Compact 10
using the Clough42 system but I’m curious to see other similar
alternatives.

thanks,
alex

I think this is the origin. It’s in Russian. I used an automatic
translator for my browser to read the thread. The one I use supplies a
funnier name than we’re using for it.

Does anyone know Oleg’s last name?
I was attracted to it because I’m making one for a wood lathe where
I’ll just rotate the spindle by hand, plus it seemed easier to change
around than other designs.

On Mon, Feb 10, 2020 at 08:26 AM, alexphredorg wrote:

Is there a GitHub or other page for the RELS project being
discussed on this list? ?I think I’ve read all of the posts but
never found a summary page or link to the source code.

I’m currently adding an electronic leadscrew to my Emco Compact 10
using the Clough42 system but I’m curious to see other similar
alternatives.

thanks,
alex

Is there a GitHub or other page for the RELS project being discussed on this list? ?I think I’ve read all of the posts but never found a summary page or link to the source code. ?

I’m currently adding an electronic leadscrew to my Emco Compact 10 using the Clough42 system but I’m curious to see other similar alternatives.?

thanks,
alex

If you just need the stepper to rotate the D head the you can easily
check and see what actual torque is needed some sort of arm with a
weight on the end keep increasing the radius until it rotates the
calculate the oz in or Nm. If you intend rotating whilst cutting that
will obviously require more torque.
Hobbing on the mill is possible but not with your existing dividing
head. You will need to rotate the component much faster. Just think
about what speed you would like to run the cutter at then divide by the
gear ratio you are hobbing. Your stepper would be really spinning when
the component is driven at 40 to 1.
The 3" is not a problem I would have thought, just jack it up on
spacers. I have hobbed using Linux CNC and also built an interface to
allow hobbing without the computer just using an Arduino Mega and
putting an encoder on the spindle. Best option if you need to make a
worm and wheel.
To hob you? obviously need to be able to tilt the head!
Richard

On 05/02/2020 11:36, Clark Panaccione via Groups.Io wrote:

Thanks, John - was that on the rotary table, or on your belt drive one?
This is the highest rated Nema 17 I found, 2.3a should work with the
little TB6560 driver recommended in the “Arduino rotary table for
dummies” thread on homemodelenginemachinist.com:
?

Any insight on the control interface? It looks like yours had
pushbuttons arranged kind of joystick-like - is that for setting up as
well as operating? I thought maybe just a couple big buttons for left
and right, in addition to a keypad and display, would make it a little
simpler for me to run.
It would be interesting to rig up something like the RELS on my little
antique horizontal universal mill for hobbing but I think the 3” swing
on my dividing head is too small to worry about powering it for that.
Don’t know how many true axes that would be, with the table running at
funny angles ; )

On Mon, Feb 3, 2020 at 10:38 PM, John Lindo wrote:

I used a Nema 17 for many years, it worked vert well, changed to
Nema 23 during the full mill conversion to CNC , thus making it a
true 4th axis machine,

--
John

Is there a GitHub or other page for the RELS project being discussed
on this list? ?I think I’ve read all of the posts but never found a
summary page or link to the source code.

I’m currently adding an electronic leadscrew to my Emco Compact 10
using the Clough42 system but I’m curious to see other similar
alternatives.

thanks,
alex

Is there a GitHub or other page for the RELS project being discussed
on this list? ?I think I’ve read all of the posts but never found a
summary page or link to the source code.

I’m currently adding an electronic leadscrew to my Emco Compact 10
using the Clough42 system but I’m curious to see other similar
alternatives.

thanks,
alex

Is there a GitHub or other page for the RELS project being discussed
on this list? ?I think I’ve read all of the posts but never found a
summary page or link to the source code.

I’m currently adding an electronic leadscrew to my Emco Compact 10
using the Clough42 system but I’m curious to see other similar
alternatives.

thanks,
alex

> -----Original Message-----

> From: [email protected] [mailto:[email protected]] On

> Behalf Of Richard

> Sent: February-07-20 11:22 AM

> To: [email protected]

> Subject: Re: [digitalhobbyist] Stepper size for small dividing head? #ESI

>

> Those calculation are why I modified my home made A axis head for the

> mill from about 40:1 worm drive to 1:1 direct drive. I also hob at a

> rate higher than 200rpm.

> I feel an A axis should be between 1:1 and about 6:1. The other thing to

> consider is that there is a working load on the stepper whilst Hobbing

> as opposed to just being locked in an incremented position whilst

> machining takes place.

> ATB

> Richard

> On 07/02/2020 15:27, Clark Panaccione via Groups.Io wrote:

> > Thanks, John,

> > Yes, measuring how much torque it actually takes is a good idea!

> >

> > I think the lowest speed on the Stark mill, which doesn?t have a back

> > gear, is around 200rpm, so a 20-tooth gear would have to turn 10rpm to

> > keep up and with the 40:1 DH the stepper has to go 400rpm? According

> > to the chart on the page I linked before the 111oz-in stepper torque?s

> > about 70% stall, or 80% max running torque at that speed - does that

> > seem right?

> > But I think a bigger worm gear would be better for that - this

> > dividing head?s the smallest one I have!

> >

> > Cutting cams like it shows here is another intriguing use of a DH, and

> > would use its tilting capability, but at this point that?s probably

> > better done with cnc.

> >

>

> >

> > On Wed, Feb 5, 2020 at 08:12 AM, Richard wrote:

> >

> >???? If you just need the stepper to rotate the D head the you can easily

> >???? check and see what actual torque is needed some sort of arm with a

> >???? weight on the end keep increasing the radius until it rotates the

> >???? calculate the oz in or Nm. If you intend rotating whilst cutting that

> >???? will obviously require more torque.

> >???? Hobbing on the mill is possible but not with your existing dividing

> >???? head. You will need to rotate the component much faster. Just think

> >???? about what speed you would like to run the cutter at then divide

> >???? by the

> >???? gear ratio you are hobbing. Your stepper would be really spinning when

> >???? the component is driven at 40 to 1.

> >???? The 3" is not a problem I would have thought, just jack it up on

> >???? spacers. I have hobbed using Linux CNC and also built an interface to

> >???? allow hobbing without the computer just using an Arduino Mega and

> >???? putting an encoder on the spindle. Best option if you need to make a

> >???? worm and wheel.

> >???? To hob you? obviously need to be able to tilt the head!

> >???? Richard

> >

> >???? On 05/02/2020 11:36, Clark Panaccione via Groups.Io wrote:

> >

> >?????? ??Thanks, John - was that on the rotary table, or on your belt

> >???????? drive one?

> >???????? This is the highest rated Nema 17 I found, 2.3a should work

> >???????? with the

> >???????? little TB6560 driver recommended in the ?Arduino rotary table for

> >???????? dummies? thread on homemodelenginemachinist.com:

> >????????

> >?? ??????

> >???????? Any insight on the control interface? It looks like yours had

> >???????? pushbuttons arranged kind of joystick-like - is that for

> >???????? setting up as

> >???????? well as operating? I thought maybe just a couple big buttons

> >???????? for left

> >???????? and right, in addition to a keypad and display, would make it

> >???????? a little

> >???????? simpler for me to run.

> >???????? It would be interesting to rig up something like the RELS on

> >???????? my little

> >???????? antique horizontal universal mill for hobbing but I think the

> >???????? 3? swing

> >???????? on my dividing head is too small to worry about powering it

> >???????? for that.

> >???????? Don?t know how many true axes that would be, with the table

> >???????? running at

> >???????? funny angles ; )

> >

> >???????? On Mon, Feb 3, 2020 at 10:38 PM, John Lindo wrote:

> >

> >???????? I used a Nema 17 for many years, it worked vert well, changed to

> >???????? Nema 23 during the full mill conversion to CNC , thus making it a

> >???????? true 4th axis machine,

> >

> >???????? --

> >???????? John

> >

> >

>

>

>

Thanks, John,
Yes, measuring how much torque it actually takes is a good idea!

I think the lowest speed on the Stark mill, which doesn’t have a back
gear, is around 200rpm, so a 20-tooth gear would have to turn 10rpm to
keep up and with the 40:1 DH the stepper has to go 400rpm? According
to the chart on the page I linked before the 111oz-in stepper torque’s
about 70% stall, or 80% max running torque at that speed - does that
seem right?
But I think a bigger worm gear would be better for that - this
dividing head’s the smallest one I have!

Cutting cams like it shows here is another intriguing use of a DH, and
would use its tilting capability, but at this point that’s probably
better done with cnc.


On Wed, Feb 5, 2020 at 08:12 AM, Richard wrote:

If you just need the stepper to rotate the D head the you can easily
check and see what actual torque is needed some sort of arm with a
weight on the end keep increasing the radius until it rotates the
calculate the oz in or Nm. If you intend rotating whilst cutting that
will obviously require more torque.
Hobbing on the mill is possible but not with your existing dividing
head. You will need to rotate the component much faster. Just think
about what speed you would like to run the cutter at then divide
by the
gear ratio you are hobbing. Your stepper would be really spinning when
the component is driven at 40 to 1.
The 3" is not a problem I would have thought, just jack it up on
spacers. I have hobbed using Linux CNC and also built an interface to
allow hobbing without the computer just using an Arduino Mega and
putting an encoder on the spindle. Best option if you need to make a
worm and wheel.
To hob you? obviously need to be able to tilt the head!
Richard

On 05/02/2020 11:36, Clark Panaccione via Groups.Io wrote:

Thanks, John - was that on the rotary table, or on your belt
drive one?
This is the highest rated Nema 17 I found, 2.3a should work
with the
little TB6560 driver recommended in the “Arduino rotary table for
dummies” thread on homemodelenginemachinist.com:


Any insight on the control interface? It looks like yours had
pushbuttons arranged kind of joystick-like - is that for
setting up as
well as operating? I thought maybe just a couple big buttons
for left
and right, in addition to a keypad and display, would make it
a little
simpler for me to run.
It would be interesting to rig up something like the RELS on
my little
antique horizontal universal mill for hobbing but I think the
3” swing
on my dividing head is too small to worry about powering it
for that.
Don’t know how many true axes that would be, with the table
running at
funny angles ; )

On Mon, Feb 3, 2020 at 10:38 PM, John Lindo wrote:

I used a Nema 17 for many years, it worked vert well, changed to
Nema 23 during the full mill conversion to CNC , thus making it a
true 4th axis machine,

--
John