On Sun, 19 Mar 2017 11:23:09 -0700, gyansorova wrote:

> On Saturday, March 18, 2017 at 11:08:17 PM UTC+13, Tauno Voipio wrote:
>> On 17.3.17 23:13, Tim Wescott wrote:
>> > On Fri, 17 Mar 2017 21:13:26 +0200, Tauno Voipio wrote:
>> >
>> >> On 17.3.17 21:05, Tim Wescott wrote:
>> >>> On Thu, 16 Mar 2017 23:03:20 -0700, robert bristow-johnson wrote:
>> >>>
>> >>>> On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott
>> >>>> wrote:
>> >>>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>> >>>>>
>> >>>>>> I saw a real fast one armed robot. I was wondering if stepper
>> >>>>>> motors was used or closed-loop control. I know NC machines use
>> >>>>>> stepper motors but I heard the torque isn't as good as dc
>> >>>>>> motors.
>> >>>>>
>> >>>>> * Stepper motors are bigger for the torque that they'll deliver,
>> >>>>> and getting the most torque out of them requires trickery.  Well
>> >>>>> known and established trickery, but still trickery.
>> >>>>
>> >>>> what's the trickery?  like quadrature sinusoids applied to the two
>> >>>> windings with frequency/phase synchronized to the movement of the
>> >>>> rotor?
>> >>>>
>> >>>> otherwise, i can't imagine what you can do.
>> >>>
>> >>> Running them at their rated current, even if it means running them
>> >>> above their rated voltage (so torque doesn't go down with speed).
>> >>>
>> >>> Ramping the speed up and down so that the torque required to
>> >>> accelerate the armature doesn't exceed the maximum.
>> >>>
>> >>> Microstepping may help, to -- I dunno about that one, but I'm
>> >>> willing to believe it.
>> >>>
>> >>> I'm not the world's foremost expert on steppers -- I usually try to
>> >>> deal with them by not using them.  So others will probably give
>> >>> better,
>> >>> or at least more thorough, answers.
>> >>
>> >>
>> >> Microstepping helps, but without position feedback you still need to
>> >> observe the acceleration and deceleration limitations.
>> >>
>> >> Been there - done that. It was not fun with the tools of the 1980's.
>> >
>> > My two worst experiences with motors involved mechanical engineers
>> > that selected stepper motors incorrectly without review by EEs or
>> > Software Engineering, and then dumped them into my lap to make it
>> > good.
>> >
>> > It's a sad thing when a software guy has to explain to a mechanical
>> > guy that it takes torque to accelerate a flywheel.
>> >
>> > On one I managed to pull the fat from the fire.  On the other we
>> > never did get the system to reliably meet spec, although we did make
>> > it much better in the end.
>> 
>> 
>> Right. It may be difficult to convince the mech people that the worst
>> load to a stepper is one with hefty inertia.
>> 
>> --
>> 
>> -TV
> 
> So I think you guys are saying that steppers are ok when the load is
> light and speed is not an issue. If you want high torque and speed then
> go for servos.

More or less.  There's always exceptions -- you can sense the position of 
a stepper motor and treat the assembly as a brushless motor, but that 
gets as complicated as using a servo, so why bother?

If you want high torque and speed _for the size of the motor_ then get a 
servo.

> They do have steppers for NC machines of course but they
> don't have to go that fast I suppose.

I believe that's the tradeoff, yes.  Or they can use honkin' big power-
hungry motors and it doesn't matter much.  I've never built an NC 
machine, so take what I say with the right-sized grain of salt.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

I'm looking for work -- see my website!

On Saturday, March 18, 2017 at 11:08:17 PM UTC+13, Tauno Voipio wrote:
> On 17.3.17 23:13, Tim Wescott wrote:
> > On Fri, 17 Mar 2017 21:13:26 +0200, Tauno Voipio wrote:
> >
> >> On 17.3.17 21:05, Tim Wescott wrote:
> >>> On Thu, 16 Mar 2017 23:03:20 -0700, robert bristow-johnson wrote:
> >>>
> >>>> On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott wrote:
> >>>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
> >>>>>
> >>>>>> I saw a real fast one armed robot. I was wondering if stepper motors
> >>>>>> was used or closed-loop control. I know NC machines use stepper
> >>>>>> motors but I heard the torque isn't as good as dc motors.
> >>>>>
> >>>>> * Stepper motors are bigger for the torque that they'll deliver, and
> >>>>> getting the most torque out of them requires trickery.  Well known
> >>>>> and established trickery, but still trickery.
> >>>>
> >>>> what's the trickery?  like quadrature sinusoids applied to the two
> >>>> windings with frequency/phase synchronized to the movement of the
> >>>> rotor?
> >>>>
> >>>> otherwise, i can't imagine what you can do.
> >>>
> >>> Running them at their rated current, even if it means running them
> >>> above their rated voltage (so torque doesn't go down with speed).
> >>>
> >>> Ramping the speed up and down so that the torque required to accelerate
> >>> the armature doesn't exceed the maximum.
> >>>
> >>> Microstepping may help, to -- I dunno about that one, but I'm willing
> >>> to believe it.
> >>>
> >>> I'm not the world's foremost expert on steppers -- I usually try to
> >>> deal with them by not using them.  So others will probably give better,
> >>> or at least more thorough, answers.
> >>
> >>
> >> Microstepping helps, but without position feedback you still need to
> >> observe the acceleration and deceleration limitations.
> >>
> >> Been there - done that. It was not fun with the tools of the 1980's.
> >
> > My two worst experiences with motors involved mechanical engineers that
> > selected stepper motors incorrectly without review by EEs or Software
> > Engineering, and then dumped them into my lap to make it good.
> >
> > It's a sad thing when a software guy has to explain to a mechanical guy
> > that it takes torque to accelerate a flywheel.
> >
> > On one I managed to pull the fat from the fire.  On the other we never
> > did get the system to reliably meet spec, although we did make it much
> > better in the end.
> 
> 
> Right. It may be difficult to convince the mech people that
> the worst load to a stepper is one with hefty inertia.
> 
> -- 
> 
> -TV

So I think you guys are saying that steppers are ok when the load is light and speed is not an issue. If you want high torque and speed then go for servos. They do have steppers for NC machines of course but they don't have to go that fast I suppose.

On 17.3.17 23:13, Tim Wescott wrote:
> On Fri, 17 Mar 2017 21:13:26 +0200, Tauno Voipio wrote:
>
>> On 17.3.17 21:05, Tim Wescott wrote:
>>> On Thu, 16 Mar 2017 23:03:20 -0700, robert bristow-johnson wrote:
>>>
>>>> On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott wrote:
>>>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>>>>>
>>>>>> I saw a real fast one armed robot. I was wondering if stepper motors
>>>>>> was used or closed-loop control. I know NC machines use stepper
>>>>>> motors but I heard the torque isn't as good as dc motors.
>>>>>
>>>>> * Stepper motors are bigger for the torque that they'll deliver, and
>>>>> getting the most torque out of them requires trickery.  Well known
>>>>> and established trickery, but still trickery.
>>>>
>>>> what's the trickery?  like quadrature sinusoids applied to the two
>>>> windings with frequency/phase synchronized to the movement of the
>>>> rotor?
>>>>
>>>> otherwise, i can't imagine what you can do.
>>>
>>> Running them at their rated current, even if it means running them
>>> above their rated voltage (so torque doesn't go down with speed).
>>>
>>> Ramping the speed up and down so that the torque required to accelerate
>>> the armature doesn't exceed the maximum.
>>>
>>> Microstepping may help, to -- I dunno about that one, but I'm willing
>>> to believe it.
>>>
>>> I'm not the world's foremost expert on steppers -- I usually try to
>>> deal with them by not using them.  So others will probably give better,
>>> or at least more thorough, answers.
>>
>>
>> Microstepping helps, but without position feedback you still need to
>> observe the acceleration and deceleration limitations.
>>
>> Been there - done that. It was not fun with the tools of the 1980's.
>
> My two worst experiences with motors involved mechanical engineers that
> selected stepper motors incorrectly without review by EEs or Software
> Engineering, and then dumped them into my lap to make it good.
>
> It's a sad thing when a software guy has to explain to a mechanical guy
> that it takes torque to accelerate a flywheel.
>
> On one I managed to pull the fat from the fire.  On the other we never
> did get the system to reliably meet spec, although we did make it much
> better in the end.


Right. It may be difficult to convince the mech people that
the worst load to a stepper is one with hefty inertia.

-- 

-TV

On Fri, 17 Mar 2017 21:13:26 +0200, Tauno Voipio wrote:

> On 17.3.17 21:05, Tim Wescott wrote:
>> On Thu, 16 Mar 2017 23:03:20 -0700, robert bristow-johnson wrote:
>>
>>> On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott wrote:
>>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>>>>
>>>>> I saw a real fast one armed robot. I was wondering if stepper motors
>>>>> was used or closed-loop control. I know NC machines use stepper
>>>>> motors but I heard the torque isn't as good as dc motors.
>>>>
>>>> * Stepper motors are bigger for the torque that they'll deliver, and
>>>> getting the most torque out of them requires trickery.  Well known
>>>> and established trickery, but still trickery.
>>>
>>> what's the trickery?  like quadrature sinusoids applied to the two
>>> windings with frequency/phase synchronized to the movement of the
>>> rotor?
>>>
>>> otherwise, i can't imagine what you can do.
>>
>> Running them at their rated current, even if it means running them
>> above their rated voltage (so torque doesn't go down with speed).
>>
>> Ramping the speed up and down so that the torque required to accelerate
>> the armature doesn't exceed the maximum.
>>
>> Microstepping may help, to -- I dunno about that one, but I'm willing
>> to believe it.
>>
>> I'm not the world's foremost expert on steppers -- I usually try to
>> deal with them by not using them.  So others will probably give better,
>> or at least more thorough, answers.
> 
> 
> Microstepping helps, but without position feedback you still need to
> observe the acceleration and deceleration limitations.
> 
> Been there - done that. It was not fun with the tools of the 1980's.

My two worst experiences with motors involved mechanical engineers that 
selected stepper motors incorrectly without review by EEs or Software 
Engineering, and then dumped them into my lap to make it good.

It's a sad thing when a software guy has to explain to a mechanical guy 
that it takes torque to accelerate a flywheel.

On one I managed to pull the fat from the fire.  On the other we never 
did get the system to reliably meet spec, although we did make it much 
better in the end.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

I'm looking for work -- see my website!

On 17.3.17 21:05, Tim Wescott wrote:
> On Thu, 16 Mar 2017 23:03:20 -0700, robert bristow-johnson wrote:
>
>> On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott wrote:
>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>>>
>>>> I saw a real fast one armed robot. I was wondering if stepper motors
>>>> was used or closed-loop control. I know NC machines use stepper
>>>> motors but I heard the torque isn't as good as dc motors.
>>>
>>> * Stepper motors are bigger for the torque that they'll deliver, and
>>> getting the most torque out of them requires trickery.  Well known and
>>> established trickery, but still trickery.
>>
>> what's the trickery?  like quadrature sinusoids applied to the two
>> windings with frequency/phase synchronized to the movement of the rotor?
>>
>> otherwise, i can't imagine what you can do.
>
> Running them at their rated current, even if it means running them above
> their rated voltage (so torque doesn't go down with speed).
>
> Ramping the speed up and down so that the torque required to accelerate
> the armature doesn't exceed the maximum.
>
> Microstepping may help, to -- I dunno about that one, but I'm willing to
> believe it.
>
> I'm not the world's foremost expert on steppers -- I usually try to deal
> with them by not using them.  So others will probably give better, or at
> least more thorough, answers.


Microstepping helps, but without position feedback you still need to
observe the acceleration and deceleration limitations.

Been there - done that. It was not fun with the tools of the 1980's.

-- 

-TV

On Thu, 16 Mar 2017 23:03:20 -0700, robert bristow-johnson wrote:

> On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott wrote:
>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>> 
>> > I saw a real fast one armed robot. I was wondering if stepper motors
>> > was used or closed-loop control. I know NC machines use stepper
>> > motors but I heard the torque isn't as good as dc motors.
>> 
>> * Stepper motors are bigger for the torque that they'll deliver, and
>> getting the most torque out of them requires trickery.  Well known and
>> established trickery, but still trickery.
> 
> what's the trickery?  like quadrature sinusoids applied to the two
> windings with frequency/phase synchronized to the movement of the rotor?
> 
> otherwise, i can't imagine what you can do.

Running them at their rated current, even if it means running them above 
their rated voltage (so torque doesn't go down with speed).

Ramping the speed up and down so that the torque required to accelerate 
the armature doesn't exceed the maximum.

Microstepping may help, to -- I dunno about that one, but I'm willing to 
believe it.

I'm not the world's foremost expert on steppers -- I usually try to deal 
with them by not using them.  So others will probably give better, or at 
least more thorough, answers.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

I'm looking for work -- see my website!

On Thu, 16 Mar 2017 20:26:13 +0200, Tauno Voipio wrote:

> On 16.3.17 17:20, Tim Wescott wrote:
>> On Wed, 15 Mar 2017 20:55:37 -0700, gyansorova wrote:
>>
>>> On Thursday, March 16, 2017 at 7:09:09 AM UTC+13, Tim Wescott wrote:
>>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>>>>
>>>>> I saw a real fast one armed robot. I was wondering if stepper motors
>>>>> was used or closed-loop control. I know NC machines use stepper
>>>>> motors but I heard the torque isn't as good as dc motors.
>>>>
>>>> * Stepper motors are bigger for the torque that they'll deliver, and
>>>> getting the most torque out of them requires trickery.  Well known
>>>> and established trickery, but still trickery.
>>>>
>>>> * _Some_ NC machines use steppers, some use servo motors.  For a time
>>>> (and possibly still, I haven't been paying attention recently), most
>>>> of the DIY NC machine software was targeted to stepper outputs. 
>>>> However, it is very common to have a servo motor controller that
>>>> takes the same "direction + step" commands that a stepper motor
>>>> driver does.
>>>>
>>>> If you want maximum speed for the motor size, then you want a servo
>>>> motor.
>>>>
>>>> --
>>>>
>>>> Tim Wescott Wescott Design Services http://www.wescottdesign.com
>>>>
>>>> I'm looking for work -- see my website!
>>>
>>> Do position servos work 360 degrees nowadays. In the old days we had
>>> pots that limited the travel though ac servos using synchros could go
>>> right round
>>
>> If you're talking RC servos, yes, they're still limited in travel.
>>
>> If you're talking industrial servo motors, most of them come with shaft
>> encoders.  As long as your mechanism can stand doing a "home" cycle on
>> startup, you can get as many turns out of them as you wish.
>>
>> (AFAIK, the current trend is servo motors with built-in drives, so
>> you're not even driving the motor externally -- you're just supplying
>> it with power and commands.  But I'm usually involved in things that
>> are more deeply embedded than that, at least at the moment.)
> 
> 
> Before the shaft encoders, there were sin/cos potentiometers and
> resolvers (rotating I/Q transformers), also synchros, which were
> three-phase versions of the resolvers.

Yup.  Those work great.  They're still used where a homing cycle on 
startup is out of the question.

-- 
Tim Wescott
Control systems, embedded software and circuit design
I'm looking for work!  See my website if you're interested
http://www.wescottdesign.com

On 16.3.17 21:49, gyansorova@gmail.com wrote:
> On Friday, March 17, 2017 at 7:26:19 AM UTC+13, Tauno Voipio wrote:
>> On 16.3.17 17:20, Tim Wescott wrote:
>>> On Wed, 15 Mar 2017 20:55:37 -0700, gyansorova wrote:
>>>
>>>> On Thursday, March 16, 2017 at 7:09:09 AM UTC+13, Tim Wescott wrote:
>>>>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
>>>>>
>>>>>> I saw a real fast one armed robot. I was wondering if stepper motors
>>>>>> was used or closed-loop control. I know NC machines use stepper
>>>>>> motors but I heard the torque isn't as good as dc motors.
>>>>>
>>>>> * Stepper motors are bigger for the torque that they'll deliver, and
>>>>> getting the most torque out of them requires trickery.  Well known and
>>>>> established trickery, but still trickery.
>>>>>
>>>>> * _Some_ NC machines use steppers, some use servo motors.  For a time
>>>>> (and possibly still, I haven't been paying attention recently), most of
>>>>> the DIY NC machine software was targeted to stepper outputs.  However,
>>>>> it is very common to have a servo motor controller that takes the same
>>>>> "direction + step" commands that a stepper motor driver does.
>>>>>
>>>>> If you want maximum speed for the motor size, then you want a servo
>>>>> motor.
>>>>>
>>>>> --
>>>>>
>>>>> Tim Wescott Wescott Design Services http://www.wescottdesign.com
>>>>>
>>>>> I'm looking for work -- see my website!
>>>>
>>>> Do position servos work 360 degrees nowadays. In the old days we had
>>>> pots that limited the travel though ac servos using synchros could go
>>>> right round
>>>
>>> If you're talking RC servos, yes, they're still limited in travel.
>>>
>>> If you're talking industrial servo motors, most of them come with shaft
>>> encoders.  As long as your mechanism can stand doing a "home" cycle on
>>> startup, you can get as many turns out of them as you wish.
>>>
>>> (AFAIK, the current trend is servo motors with built-in drives, so you're
>>> not even driving the motor externally -- you're just supplying it with
>>> power and commands.  But I'm usually involved in things that are more
>>> deeply embedded than that, at least at the moment.)
>>
>>
>> Before the shaft encoders, there were sin/cos potentiometers and
>> resolvers (rotating I/Q transformers), also synchros, which were
>> three-phase versions of the resolvers.
>>
>> --
>>
>> -TV
>
> Yes, the military used a lot of them, maybe for gun turrets and the like, you can still buy them.


Not only military, also heavy civilian use.

-- 

-TV

On Wednesday, March 15, 2017 at 2:09:09 PM UTC-4, Tim Wescott wrote:
> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
> 
> > I saw a real fast one armed robot. I was wondering if stepper motors was
> > used or closed-loop control. I know NC machines use stepper motors but I
> > heard the torque isn't as good as dc motors.
> 
> * Stepper motors are bigger for the torque that they'll deliver, and 
> getting the most torque out of them requires trickery.  Well known and 
> established trickery, but still trickery.

what's the trickery?  like quadrature sinusoids applied to the two windings with frequency/phase synchronized to the movement of the rotor?

otherwise, i can't imagine what you can do.

r b-j

On Friday, March 17, 2017 at 7:26:19 AM UTC+13, Tauno Voipio wrote:
> On 16.3.17 17:20, Tim Wescott wrote:
> > On Wed, 15 Mar 2017 20:55:37 -0700, gyansorova wrote:
> >
> >> On Thursday, March 16, 2017 at 7:09:09 AM UTC+13, Tim Wescott wrote:
> >>> On Wed, 15 Mar 2017 10:49:13 -0700, gyansorova wrote:
> >>>
> >>>> I saw a real fast one armed robot. I was wondering if stepper motors
> >>>> was used or closed-loop control. I know NC machines use stepper
> >>>> motors but I heard the torque isn't as good as dc motors.
> >>>
> >>> * Stepper motors are bigger for the torque that they'll deliver, and
> >>> getting the most torque out of them requires trickery.  Well known and
> >>> established trickery, but still trickery.
> >>>
> >>> * _Some_ NC machines use steppers, some use servo motors.  For a time
> >>> (and possibly still, I haven't been paying attention recently), most of
> >>> the DIY NC machine software was targeted to stepper outputs.  However,
> >>> it is very common to have a servo motor controller that takes the same
> >>> "direction + step" commands that a stepper motor driver does.
> >>>
> >>> If you want maximum speed for the motor size, then you want a servo
> >>> motor.
> >>>
> >>> --
> >>>
> >>> Tim Wescott Wescott Design Services http://www.wescottdesign.com
> >>>
> >>> I'm looking for work -- see my website!
> >>
> >> Do position servos work 360 degrees nowadays. In the old days we had
> >> pots that limited the travel though ac servos using synchros could go
> >> right round
> >
> > If you're talking RC servos, yes, they're still limited in travel.
> >
> > If you're talking industrial servo motors, most of them come with shaft
> > encoders.  As long as your mechanism can stand doing a "home" cycle on
> > startup, you can get as many turns out of them as you wish.
> >
> > (AFAIK, the current trend is servo motors with built-in drives, so you're
> > not even driving the motor externally -- you're just supplying it with
> > power and commands.  But I'm usually involved in things that are more
> > deeply embedded than that, at least at the moment.)
> 
> 
> Before the shaft encoders, there were sin/cos potentiometers and
> resolvers (rotating I/Q transformers), also synchros, which were
> three-phase versions of the resolvers.
> 
> -- 
> 
> -TV

Yes, the military used a lot of them, maybe for gun turrets and the like, you can still buy them.