[Pax7]

Hello All,

I am currently developing a servo motor control prototype where I need a low voltage (around 24 VDC) brushless AC servo motor to test with.

Wanted specs of the motor:

* BLAC motor, aka PMSM (sinusoidal drive of the phases = no torque ripple)
* Max ca. 30VDC drive phase voltage
* Quadrature encoder feedback, with index signal

I have spent several hours searching the Internet for such a motor, but have not been able to find any

So, if anybody here knows where I could find such a motor, I would be very grateful!

Thanks!
/Fredrik

[Al_The_Man]

You shouldn't need to restrict the search for sinusoidal, a BLDC should work for you just as well.
Picmicro in there demo kits specify either a Pittman or they supply one direct as used in their BLDC Demo board, low voltage motors.
http://ww1.microchip.com/downloads/e...tes/00857a.pdf
Product Search - microchipDIRECT
I don't think you will find many motors of any size in that voltage range?
What method of commutation were you thinking of?
Al.

[Pax7]

Hello Al,

and thanks for your reply!

My application is a bit unusual, as it cannot have any torque ripple, plus it needs high resolution shaft positioning. So, a BLDC motor will not work as it has too much torque ripple. Also, I cannot use Hall encoders as they do not give exact shaft position information.

The prototype board I work with only has a quadrature interface, so I cannot use resolver type feedback (whose output needs to be run through a ADC).

So, I arrive at the specs in my original post. I will use three phase sinusoidal commutation.

Any more tips are very welcome!

Thanks,
Fredrik

[Al_The_Man]

I did not mean use BLDC as the method of control, rather use an available BLDC motor for your sinusoidal method, the motors are relatively identical, apart from the commutation methods.
BTW, I use BLDC motors with a Dsp centered PID controller, Galil Motion cards, and if you were to observe the smoothness or lack of cogging at very low RPM, you would not be able to differentiate between this motor and any other type of servo motor.
Al.

[Pax7]

Interesting; what I know is that BLDC motors have trapezoidal back-EMF due to their stator configuration, and are commutated using the six-step method. Is it a trapezoidal back-EMF BLDC you are running sinusoidaly you say?

[Al_The_Man]

The only difference AFAIK between the construction of the two types is the rotor magnets on SOME AC sinusoidal, they claim to have dome shaped magnet tops, otherwise they have 3 phase, star connected stator windings, I have used both in either configuration?
Just the commutation and form of the power applied is different?
Al.

[Pax7]

I have also thought about what exactly is different in the construction between the two types of motors. Here is e.g. what Microchip writes in app note AN1017:

"The main disadvantage of BLDC motors, when low
torque ripple and quieter operation are required, is the
non-sinusoidal distribution of the stator windings.
BLDC motors with non-sinusoidal winding distribution
generate trapezoidal back-EMF[...]"

I have not seen any info though on what exactly differs between non-sinusoidal distribution of the stator windings vs. sinusoidal distribution. That would be interesting to see!

[Al_The_Man]

As far as the performance, cogging and noise, as I mentioned in using a BLDC motor with Galil Motion, I see no perceptible difference between the performance of BLDC and DC brushed particularly at very low rpm where these effect are purported to appear.
If they are ran by a simple analogue signal into the transconductance amplifier drive with no PID feedback/control then BLDC do exhibit these traits, somewhat.
Al.

[mike_Kilroy]

Interesting; what I know is that BLDC motors have trapezoidal back-EMF due to their stator configuration, and are commutated using the six-step method. Is it a trapezoidal back-EMF BLDC you are running sinusoidaly you say?

My take is that for 89.9% of applications you can get by just fine with a trap designed motor on a sine drive or vice versa. For that other 10.% that cannot accept the additional torque ripple, you cannot. I will swag the additional torque ripple from memory as about 3% if you mismatch like this.

If an additional 3% torque ripple on top of the matched 'good' motor design with skewed magnets of what.... swag again - its been some years since I needed to get into torque ripple measurements - say 3-5% is not a problem, a-ok. You should be able to search around http://wiki.danaher.communicode.de/t...page=Startpage website and find white papers telling you these torque ripple values by mismatching the drive and motor type as well as other torque ripple values for various of the motors they sell that will run on 24v bus as you wish.

You also can buy a more expensive motor that minimizes torque ripple various ways and then pick a drive that matches. For instance, Danaher's Home | Portescap | English motors are designed as "0 cogging" motors. they have no steel in their armatures so no magnetic attraction to pole pieces. Then Kollmorgen's PMI totally zero cogging DC motors Servo Disc | Kollmorgen are a choice. Being brush type they like the portescap are happy to run off a 24v bus.

Pick a good matching high bandwidth drive so it can correct for any variations in speed to match. I'd suggest you use a drive that accepts sine encoder for 2-3x bandwidth and performance of simple TTL encoder drive, and then let its encoder equiv output give you the ttl encoder sig of your choice from that as buffered position sig for your position controller like Galil.

[Al_The_Man]

I am assuming as you are looking for something in the 24vdc range it is for a portable application?
Unless you intend operating in a close to direct drive without any gearing, any torque ripple or cogging effect may be a moot point with BLDC?
Al.

[mike_Kilroy]

agreed!

[Pax7]

Thanks guys for the info!

I am really intrigued by these Kollmorgen servo discs! Ideally, the servo for my application shall have zero cogging, high acceleration capabilities and high torque (also continuous). And, yes, it will be a direct drive system; no gear boxes etc.
So, these discs seem close to ideal! I guess the back side is they are very expensive..? Do you have a ball park figure for the price range?

The 24V requirement is mostly for my prototype system, which I work with now. When I have all MCU programming done (commutation, feedback, USB comms etc.), I will change the servo drive stage electronics from 30V max now and (probably) go for a 230V servo instead.

[mike_Kilroy]

u9 about $ 1000; u12 about $ 1500.

[Iron-Man]

Hello All,

I am currently developing a servo motor control prototype where I need a low voltage (around 24 VDC) brushless AC servo motor to test with.

Wanted specs of the motor:

* BLAC motor, aka PMSM (sinusoidal drive of the phases = no torque ripple)
* Max ca. 30VDC drive phase voltage
* Quadrature encoder feedback, with index signal

I have spent several hours searching the Internet for such a motor, but have not been able to find any

So, if anybody here knows where I could find such a motor, I would be very grateful!

Thanks!
/Fredrik

Try the Yaskawa drives, SGMMJ etc. I have seen some of these on ebay.
See pages 28 and 29; http://www.yaskawa.com/site/dmservo....E-S800-38C.pdf

Ebay NEW YASKAWA SGMAJ -II AC SERVO MOTOR ENCODER CABLES 80W 24V servo; NEW YASKAWA SGMAJ -II AC SERVO MOTOR ENCODER CABLES 80W 24V SGMAJ-A8CAA2C | eBay

Hope this helps,

Iron-Man

[Pax7]

Try the Yaskawa drives...

Thanks Iron-Man, these actually look promising!

I looked into the specs/manuals of the motors, and what I am not clear about is the incremental encoder wiring (see pg 125, section 6.2.2 in the pdf you linked to). PS and /PS are serial signal (I found in another manual them to be serial data + and -). The question is, are those quadrature signals, or some other scheme/protocol? Do you know?

[Al_The_Man]

It would appear they have serial encoder signal from motor to the drives, and then a pseudo differential encoder system is output from the drive for use with a controller?
Al.

[inventor83]

An inexpensive alternative for a zero cogging motor would be the coreless ironless BLDC motors used for RC helicopters and the such. I do believe they are right around the 24V mark aswell.

[mike_Kilroy]

An inexpensive alternative for a zero cogging motor would be the coreless ironless BLDC motors used for RC helicopters and the such. I do believe they are right around the 24V mark aswell.

yep, without telling power or torque and speed required, no one cal help you. motors come in all sizes from rc helicopter to 1000,#-ft 100rpm rating in "24vdc with encoder" size.

You must tell what SIZE in speed and torque you want for better help.

[Pax7]

yep, without telling power or torque and speed required, no one cal help you. motors come in all sizes from rc helicopter to 1000,#-ft 100rpm rating in "24vdc with encoder" size.

You must tell what SIZE in speed and torque you want for better help.

Hello Mike,

If it those specs would have mattered much to me, I would have told you

The purpose of my project is to develop an actual servo drive system; first of all the firmware interfacing the servo drive stage power electronics (H-bridge(s)). This means I will interface the servo directly and controlling all aspects of servo phase driving, as well as reading out info on shaft position (via QEI), back EMF etc to input to my servo control loop software.

My application is somewhat special, as I do not care so much about speed, but about controlling torque. Max speed should be at least 800 rpm or so I guess though.

So, I would like a motor with little cogging and friction, and preferably AC driven, as that is the commutation style I am most interested in.

As this motor only is for a small sized prototype development system, max torque figures are less important.
Fast acceleration, low inertia and low inductance (electrical time constant) are preferred though.

Thanks,

[mike_Kilroy]

now that you gave at least ONE specification, a motor may be found. I was going to recommend a motor I recently used that was rated 17vac rms for 6000rpm rated 123amp continuous and 100 oz-in cont output for $ 12,000.00 since you had no specs to go on other than "24v."

So you would like 800rpm so your motor basic winding spec Kb = 24v/0.8krpm = 30v/krpm.

I would suggest ANY model Kollmorgen AKM11xxxxxxx - all meet that spec, are low cost (guess 200 euro), low cogging, low inertia, fast accel, readily available from local Swiss distributor.

AKM