Motion Control

[ASTDan]

Hello,

I have been quoting on a motion control job and I wanted to ask a question regarding NI's Motion control boards.

What are the pro's/con's of using a NI Motion control boad over an intelligent motor drive. An intelligent motor drive you can send commands via the serial port or control the position,velocity,or torque using an analog output. The drives I have seen you can set the PID parameters and also store moves to be called using RS232. So with all that functionality what does using a NI motion board get you?

Thanks

Dan

[Louis Manfredi]

Hi Dan:

I think it depends on the details of what you are doing... For example, I have a couple of applications where I needed simple two and three axis stepper motor controls. I ended up using Stepnet controllers from Copley Controls, controlled through a serial port to one unit which passed commands to others. Each controller has all the limit switch, home switch, acceleration, torque & speed limit I needed for my fairly simple applications, and they were a little cheaper than the NI motor control board path (not only the cost of the control board, but also separate amplifiers, as I recall.)

I did not need to control to a particular path. All I cared about was that I ended up where I wanted to be in x,y,z at each step in my process-- It didn't matter to me if I got to X before Y, or whether the path between one point and the next followed any specific curve.

Beyond avoiding mechanical resonance, overload, or alternatively wasting a lot of time getting where I was going, I didn't much care about the details of accleration or maximum velocity either.

If I'd cared about that sort of stuff, the NI board approach would have been more attractive--- I would have had to go to a more costly non-NI hardware solution, in which case the NI gear would have been cost-competitive for hardware. If the NI gear had been cost-competitive in a hardware sense, I would have chosen it to benefit from the (I suspect) more seamless integration with LabVIEW.

Just my $0.02-- be interesting to hear what others think.

Louis

[ASTDan]

The applications I have been looking at have been very simple. I have had a couple that I just needed to control a 2 axis stage using a servo motor. I just needed to go to a position. Path was not a factor.

I had another application where they needed to control Torque which was just controlling the current. Again very simple applicaiton. The drive had digital inputs for home and out of limit contitions.

So from what I gather simple motion applicaitons can be handled with an intelligent drive. Where as a complex motion you are better off with the NI Motion control card.

Is that a fair statement?

Dan

On a releated note.

What is a good reference for motion control? I ordered the Motion control Cource Kit from NI. Is there another resource? I am looking for a good book or article.

Thanks

Dan

[K_seeker]

It all depends on the type of application. For simpler (one axis or multiple axis with no co-ordinated motion) and slower applications stand alone controllers may be more appropraie. But if the your applications requires (or may eventually require) tight integration with your DAQ and IMAQ hardware than it's best to go with NI Motion. I once worked on an application where a camera was moved along an axis and the images were captured while it was moving - at every so many steps. Even though it was a simple application by Using NI Motions Breakpoint feature (http://zone.ni.com/devzone/cda/tut/p/id/4214) I was able save on wiring and control panel space, because required signals were routed through RTSI bus and no additional wiring from stand alone motion controller was required.

Another big advantage I see in using NI Motion (if you are using motion control in a Labview based test / manuafcturing system) is that it reduces my software count. This reduces programming and documentation hassles. For example if you were to use motion controller from Parker your motion control part of programming is done using Parker programming utility called Motion Planner. The Labview is most likely using ActiveX controls to interface with the motion hardware. This , in my view, makes your program less readable to a Labview programmer. Now if you need to make changes to the code you will be going back and forth between Labview and Motion Planner. By using NI only solution the software all the motion programming is done from Labview environment. In my view it's a big advantage.

[Louis Manfredi]

QUOTE(ASTDan @ Jul 31 2007, 02:24 PM)

So from what I gather simple motion applicaitons can be handled with an intelligent drive. Where as a complex motion you are better off with the NI Motion control card.

Is that a fair statement?

Fair summary of my opinion, but just my opinion. K_seeker seems to be in general agreement 'though.

By the Way K_seeker: Welcome to LAVA! :beer:

Louis

[PaulG.]

I have some experience working with MicroMo Electronics motion control products and NI Motion. With what I needed to do the MicroMo controller I had was overkill. Their motion controllers are extremely powerful with tons of functionality built right into the unit and I eventually didn't even need NI Motion. I would have a hard time imagining something so complex that I would be forced to use it. Just my $0.02.

[task]

I'm using a PCI-7352 Motion Control board to control my Parker motor (Compax3). I've used this set-up before...two motors, one driving (velocity control) and one absorbing (torque control). My issue right now is the customer's specs state that the velocity has to be a sine wave. I've been trying to modify the "One-Axis Contour Move with Position Monitor.vi" using buffers etc. but it seems the fastest update is only 10ms. I have to run the motor at 2-3 Hz going 180 degrees in both directions (durability testing). I've tried writing my own stuff but with inherent delays the motor does not run smooth. Does anybody out there have experience with using a sine-wave velocity control?

Thanks.

[task]

Finally got it to work after revisiting calculus and trig...glad my intern is a college student with this info still fresh. None of the examples provided with NI-Motion were very helpful except for the "configure buffer" one.

Thanks.

[Mike Ashe]

Glad to hear that you got your project working (aren't bright eyed interns great?). Since there is not much motion control here on LAVA (so far) it would be great if you posted a short example of what you did to solve this. Every little technique explained adds up.

Thanks!