AMC/Copely sinusoidal commutaion

[roger_e]

So the saga continues

We are trying to get sinusoidal commutaion for some motors, I prefer AC brushless.

When I talk to AMC and Copely they state the only way to get sinusoidal commutation is for that all to fall on the controller...and the sinusoidal commutation drives do not support any encoder feed-back...

Which is fine with me. How ever I have read here and from other sources that you can get sinusoidal commuation by using the Halls sensors and the shaft encoder.

We are attempting to purchase a system, and we want to use EMC2 as the controller....it is priamarily proof of concept stuff but we are allways looking at the comercial applications..

What is the deal with that?

What am I missing?

[Al_The_Man]

Maybe you are confusing AC sinusoidal with BLDC?
Also AMC maybe thought you were referring to AC sinusoidal command signal to the drive, where the controller provides two phases and the third is interpolated from these two.
If halls are used in an AC sinusoidal commutation it is just to locate the initial position of the rotor, after that the encoder is used to commutate the 3 phase, or a resolver can be used also.
This link shows the BLDC trapezoidal drive.
http://users.tinyworld.co.uk/flecc/4...otor031102.swf
A study of the different types of AMC AC sinusoidal drives will explain the different methods used.
Some accept 10v analogue, others AC sinusoidal signal.
Al.

[roger_e]

Al,

I am talking Analog Amplifiers from AMC and Copely

I was on the phone with the tech at Advanced motion controls today. This was reference a build we are doing.

I am no EE, I am a machinist and machine builder(mechanics). The electrical control stuff we outsource(purchase) now we are trying to get a handle on this because we have couple specific application machines to do, and prepackaged systems are not meeting the grade.

He stated in no uncertain terms"any B or BE series drive is trapezoidal commuataion no matter what encoder you are using...any S series drive is Sinusoidal commutaion, furthermore there are no encoder inputs on the drive/amp for the s Series...everything goes to the controller"

BTW I got the same response from Copley controls

Maybe I have my vocabulary mixed up, or maybe they do, I trust the maker of my controller, and your response about commutation I have read before several times on this forum..I just don't see why the dot's don't connect(I usually don't have a problem connecting the dots).

If there are no encoder inputs at the Amplifier then...it has to be the controller.

If the inputs are going into the Amplifier then it has to be the amplifier(and the analog controllers from copely and amc only do trapezoidal commutaion)...unless it is simply passing the signal along...right? Then at that point it would be the controller.

IF all of the above is true..then the techs at AMC/Copely are unaware of the applications mentioned on this forum....

Or I have this hopelesly jumbled

Thanks for your response

[Al_The_Man]

Depending on your application, BLDC may be just fine, (B prefix for AMC).
These use analogue command input, the position loop on just about all the AMC and Copley is principally closed by the control, the drive just requires the correct commutation in order to drive the motor.
IOW, there is no position loop from motor to drive, AMC recommend using the drive in the torque mode where the PID loop is closed via encoder to CNC or positioning controller.
What is the positioning device or software you are using?
If you already have the motors, what type are they?
Al.

[samco]

I guess it is hard to understand what you are looking for. Take the amc b or be (actually be drives can use the encoder for velocity mode) drives. These drives take an input +/-10 and depending on the mode output motor movement based on that input - either open loop, current loop, or encoder velosity.

for an emc application you would have something like this.. (lets use a mesa interface card)

Emc2->mesa pci card(5i20)->mesa servo interface card(7i33) +/-10v-> Amc drive -> servo -> Encoder -> same mesa servo interface card(7i33) -> same mesa pci card(5i20)->Emc2.
Emc closes the position loop using pid.

are you thinking of 'smart' drives that close loop within them selves? these usually take step/dir signals. Like the gecko or uhu servo drives. step and direction servo drives are evil imho

sam

[roger_e]

We are trying to use Pico system, right now and John has been very patient and informative.

We are starting with the controller Pico(PPMC2) and EMC2.

Positioning is gonna be with a renishaw RGH24 series (whatever works best with PICO) and then a shaft encoder on the motor themselves.

When I can get a consensus on the optimal setup then we will mary the AC Brushless Motors up...

The computer is bein made now specifically for this purpose

Essentially we are building from the ground up, and we have satrted with the controller(electronics/software) becasue we got a handle on mechanics and fabrication, and we can CNC our own parts.

Honestly I was hoping to get a performace system not just "a system". I make things for a living so "CNC" is not an adventure (although this topic has become very involved) and I allready have machines that cut things pretty accurately(I make optiacl components)

It may turn out that putting a system together is not as good as, or cheaper than buying one like BeckHoff, or MDSI

But I will give it a shot

[Al_The_Man]

Do you intend to use a dual feedback system with the Renishaw scale and motor encoder?
If so there is a very informative instructional video on the Galil Motion site on the ramifications and tuning of the PID loop when a dual loop is used.
Al.

[roger_e]

Samco

So maybe that is my fundamental misunderstanding: EMC2 closes the loop...is that the same as comutating the drive?

I understand the concept....my machine is open and we have traced all the wires out. We have well drawn schematics(along with quotes) from Delta-Tau, BeckHoff and Softservo.

That EMC closes the loop is the idea, but I am talking about the shape of the wave that is closing the loop(?). I have read a couple of studies that indicate the analog sinusoidal wave is going to give me the smoothest motion, and if the analog sytem is well tuned the repsonse time is better than digital.

So yes I can use the BE series with Pico or Mesa, but according to the manfacturers it will not give me sinusoidal commutation. Which is what I am looking for.

So that being said if your use of the MESA card leaves all of those functions to EMC2 (including commutation) then I will give them a call in the morning.

Or have I fallen further down the rabbit hole?

[roger_e]

Al,

Yes we are going to use dual feedback. The machine requires very tight tolerances(optics).

We are honing dovetails now for cross slides.

I looked at Galil...how ever they sugested a compnay for CNC control that I am inclined to not do business with. The first tech at Galil didn't even know what EMC2 was....and then he referenced that other company whose machines I have stumbled on across the years...never in precision situations allways plasma torches or gantry routers etc...

John Elson(Pico) has been very informative and he is only 90miles up the highway.


I will check the video tonight.

Thanks

[Al_The_Man]

So yes I can use the BE series with Pico or Mesa, but according to the manfacturers it will not give me sinusoidal commutation. Which is what I am looking for.

It appears you are confusing two areas concerning the drive, the command (motion card) does not necessarily relate to whether the motor control is sinusoidal etc, except strictly in the case of such as the Galil card which CAN output a sinusoidal command to the drive, this allows very economical design criteria in the drive itself, but they will also operate into analogue control drives and can be brushed, DCBL or AC sinusoidal when it comes to the motor control via a motion card.
All the above refers to motor drive method but not positioning.
In the case of a motion card, this is done by its positioning loop and it is not necessarily aware of the motor technology or commutation method itself.
The Galil site referred to before, has very informative instructional videos on closing servo loops and tuning, and generally apply whatever motion card is used.
BTW, Mach has a Galil plug in for true closed loop control.
Al.

[roger_e]

yes,

i am further down the rabbit hole...

I know what PID is....I can actually calculate Derivatives, Integrals and postion functions. It is calculus and the copmputer does it a gazillion times a second, me one an hour oh well....

I still don't see the disconnect between the commutation of the drives and the controller...the manfacturers of the the AMPS/Drives explicitly state that their drives must be commutated by the controller (software/hardware) if you want sinusoidal commutaion.

And I obviously do not understand your very informative reponses.

So I will leave at that and take a second look at the qutoes on my desk.

Thanks

Roger

[Al_The_Man]

I don't wish to belabour the point, but just as a further example as regards the AMC drives, the SE30A will input a ±10vdc analogue command signal and use hall effect signal from the motor for start up only and from then on use the encoder feedback for sinusoidal control of the motor.
The encoder signal can be passed through to the motion card for the PID loop.
With the S30A40C series, these only require a 2 phase sinusoidal command signal for the AC sinusoidal operation of the drive, with no other feedback device required, this is the version that is much more economical to produce.
Al.