[Larken]

A closed loop stepper does have applications for low speed high torque, but lets be honest, after 10 turns/second a stepper has very little torque compared to what it had at holding.
A brushed or brushless motor blows it away above a few hundred rpm.

Do you have any dyno plots to show Torque vs rpm ?

[raymondg]

Mariss,

I have a new conversion I am starting, and since I already own a number of G201's I want to use them to get the project started quickly and less expensively. However, I will eventually want to upgrade to 150-250 Watts/axis for this project, so I was wondering if these new drives will be able to run steppers in this power range? Would there be particular motor characteristics (i.e. amps per phase, voltage, inductance) I should look for to optimize for that power range?

For my application I really like the brushless and reductionless attributes of this approach.

Regards,
Raymond

[PEU]

yesterday I remembered this project, whats the current status?

Happy new year!


Pablo

[MrWild]

A closed loop stepper does have applications for low speed high torque, but lets be honest, after 10 turns/second a stepper has very little torque compared to what it had at holding.
A brushed or brushless motor blows it away above a few hundred rpm.

]]

I have a friend doing work for NASA. It seems hard vacuum does terrible things to all sorts of stuff. Out gassing of lubricants, hence the need for dry lubricants in moving devices, embritlement of all sorts. Rubber loses it's elasticity and so on. If you eliminate a gear train, you've reduced weight flown out of the gravity well, reduced mechanical fail points, reduced lube points, and simplified structure.

Mariss should be talking to the rocket boys with this. It blows servos out of the water for space applications.

[beliefdrive]

to Mariss, did you have final protype stepper servo driver board

b.r

[beliefdrive]

i have alpha step (vexta) it not like servo, it remember how many step lose and correct it.

[Larken]

I have a friend doing work for NASA. It seems hard vacuum does terrible things to all sorts of stuff. Out gassing of lubricants, hence the need for dry lubricants in moving devices, embritlement of all sorts. Rubber loses it's elasticity and so on. If you eliminate a gear train, you've reduced weight flown out of the gravity well, reduced mechanical fail points, reduced lube points, and simplified structure.
Mariss should be talking to the rocket boys with this. It blows servos out of the water for space applications.

I'm not saying there are no applications for a Closed loop stepper, Im just saying there is no point expecting them to have usable power above 1000 rpm, where brushed or brushless are far more powerful.
There are lots of applications where a compact high torque, low speed stepper will be great.

I'm working on one using a DSP and plan use it in our routers with a low ratio between motor and rack.

Larry K

[tivoidethuong]

Mariss
anything news?

hope step servo soon

[Tartan5]

Just a bump, Any word on this project?

Thanks!

[Mariss Freimanis]

In order of development:

First: G215
Second: GM540
Third: Brushless-DC (PMSM motor) PID servodrive
Fourth: Step motor PID servodrive

Mariss

[Tartan5]

Hello Mariss,

Is there any further progress on this? You looked like you were so close, did some show-stopper creep up in the design?

Tartan5

[joeybagadonuts]

Hello Mariss,

Is there any further progress on this? You looked like you were so close, did some show-stopper creep up in the design?

Tartan5

I was wondering the same thing.
Two years in a hot tub is a little extreme, I wouldn't even boil a bagel that long.

JoeyB

[Mariss Freimanis]

Tartan5,

No show-stopper; it's just other, more pressing projects have taken my time. In the interim, I have changed my approach to the problem. The advent of inexpensive (< $3) FPGAs allows a fully digital implementation of the Clark-Park transforms necessary for the classic field oriented control (FOC) algorithm. This is the same method used for servo control of 3-phase motors such as brushless DC (BLDC).

Curiously there is little difference between step motors and brushless DC motors, the only significant difference being pole-count (6 poles for BLDC versus 50 poles for 1.8 degree step motors). Both are considered permanent magnet synchronous motors (PMSM). For BLDC motors, current practice is to use a fast, DSP enabled microprocessor to execute the FOC algorithm about 20,000 times a second. This method is viable for 6-pole motors but not for 50-pole motors.

A step motor pole count is about 10 times higher so scaling implies the FOC algorithm must execute 10 times faster as well. This requirement is beyond the ability of most inexpensive DSPs (digital signal processors). An FPGA coded with the FOC algorithm can easily execute it 250,000 times a second using a very low 8MHz clock frequency.

Another absolute requirement for high pole-count motors (steppers) is Field Weakening. The FOC algorithm uses two command inputs; Torque Command and Field Weakening. The field weakening input for BLDC motors is often not used (made equal to zero) because the motor speed without it is already high enough in many applications (~ 3,000 RPM). This isn't the case for step motors, the maximum speed would be only 300 RPM without field weakening and greater than 3,000 RPM using field weakening. The math needed to calculate this control input is complex and requires a microprocessor. Luckily the speed required for these calculations is modest.

My approach now is to use an FPGA / microprocessor combination; the FPGA does the fast stuff while the processor does the high IQ stuff.

Mariss

[joeybagadonuts]

Tartan5,

Have you taken a look at the Leadshine closed loop hybrid servo stepper, it has been available for a over a year.

Leadshine Closed loop stepper - YouTube

JoeyB

[LGSM3]

I too would be very interested in Mariss professional opinion on the motors and drives u show. If Gecko made something like that I'd b first in line to buy it

[Mariss Freimanis]

LGSM3,

The video shows pretty much what I would expect a from a good servomotor system. It recovers from a mild position error when over torqued, it exhibits no resonances and it's properly damped in response behavior. My negative observations are the motor is a non-standard 3-phase step motor instead of a standard 2-phase 1.8 degree motor and I don't see it going faster than 300 to 500 RPM. The encoder is also a non off-the-shelf type; both spell proprietary hardware, making it either difficult to replace and/or expensive.

The motor has a rated torque of 1.2 Nm (170 in-oz) which yields only 60 Watts of mechanical power at 500 RPM. This amount of mechanical power is on par with a non-servoed 3.5A step motor driven with a 24V to 36V step motor drive. I would expect at least 100 to 150 Watts for a viable servodrive product. If this is an FOC drive, I'm guessing it doesn't employ field weakening to extend the speed range.

What's not seen is the switching topology used to drive the motor. What I expect for an acceptable servodrive is a stone-cold motor at any speed when the motor is unloaded. I obviously can't tell from the video if it is. A cold motor has the thermal reserve to accommodate heating when encountering high load and speed transients such as acceleration or momentary overloads. This behavior depends on how the windings are switched; done improperly the motor is hot all the time.

It so happens I'm back on the step servo design again. For me it's been an on-again, off-again project for nearly a decade, mostly because other more pressing projects have interfered. The time in between hasn't wasted because it's allowed me to realize the correct approach is the field oriented control one. By avoiding mistakes I hope what I deliver will be a winner.

Mariss

[The_Doctor]

Any news on the step servo? Sorry to ask, but it's been 17 months since your last update.


Thanks
Ed