What does it take to be a "Drive Expert"


Thread Starter

David Judd

Hi all,
I have been asked to put together a synopsis on what it takes to be a Drive Expert and would like the input of people who have experience in that
field. I have worked around drives, AC, DC and Servo for a while but do not consider myself an expert. Is this expertise something that can be
gained by education or is it dependent solely on experience? I am looking for what it takes to design drive systems, from drive selection to cabinet layout to electrical drawings. Any feedback would be greatly appreciated.

David Judd
Carolina Automation, Inc.

Kaufman, George


Concerning ac servo drives and motors, please visit www.MotionOnline.com and download the Handbook of AC Servo Systems. You can get a very good overview of servo system operation and a
thorough overview of application guidelines. The Handbook also includes a reference of other excellent sources of information for servo drives.

George Kaufman
Thank you for all of the valuable replies both on and off the list. Once again, the List has been a great place to tap into a large experience base.

David Judd
Carolina Automation, Inc.
> -----Original Message-----
> From: David Judd [SMTP:[email protected]]
> Ken,
> I wanted to thank you for your prompt and well thought out reply to
> my question. I am new to the controls engineer field and find it a
> great help to be able to 'pick the brains' of some of the people who
> have been doing this for so long. The automation list is a great
> place to do so. I received a number of replies with valuable
> information. I especially like the idea that you brought forth
> about using an Excel spreadsheet to compare possible drive/motor
> choices for an application. Would you be willing to expand on this,
> would it be possible to get a generic copy of what you use?
> I have written a small program for my HP48 calculator that will
> return a minimum HP rating for web handling applications
> but it is very specific.
> It would be great to get my hands on or come up with something more
> generic that could be applied to a broader application spectrum.
> Thank you again for your input and direction.
> David Judd
> Carolina Automation, Inc.
> 864.254.0050
> -----Original Message-----
> From: Ken Brown [SMTP:[email protected]]

> David,
> You ask a question that will have as many answers as there are
> people willing to stick their necks out and state their opinion . .
> . that said, here it goes:
> First, the drive system is often the bridge between two engineering
> disciplines - Mechanical and Electrical. By far the best drives
> guys I have worked with are mechanical engineers with a knack for
> understanding the electrical aspects of the drive (converter /
> inverter) technology who also posses a strong physics / dynamics
> background. (do I hear groans and hisses from the EEs?) The
> emphasis here is that these guys usually have a better idea of what
> is going on dynamically with the app.
> Since most companies divide the responsibilities for a machine
> design among two disciplines, there is often no end to the finger
> pointing that occurs when the machine will not perform due to
> inadequacies in the "integration" of the mechanical and electrical
> aspects of the system. More often than not, those responsible for
> the electrical / software aspects of the machine are hung with
> programming around deficiencies in the mechanical / electrical
> integration.
> To be a "drives expert" you need to have a good understanding of the
> mechanical dynamics of the application AND a THOROUGH understanding
> of the capabilities and quirks of the various drive technologies.
> If you don't catch anything else about this thread . . . you will
> still do well to imprint this in your brain and never forget it!
> I spent a day last week convincing a cadre of "Corporate Drives
> Experts" that a servo drive and motor would be a sub-standard choice
> for a paper converting application that required high speed
> registration. Knowing the process, the mechanical design and the
> capabilities of ACVector technology made this design choice obvious
> and an easy one to justify.
> The evaluation process can include bandwidth, mechanical size, cost,
> accuracy, availability, conformance to corporate purchasing
> standards . . . . the list goes on an on. You also bring up a good
> point about cabinet layout, this is a huge deal when dealing with a
> 700VDC bus drive that is switching currents with IGBT's. Grounding,
> PWM noise, regenerative braking, isolation of signals between drives
> and controllers . . . if you screw this up it will cause you no end
> of pain.
> From an education standpoint, the best way that I have found to get
> my arms around the various motor technologies AND the applications
> is to model them on a plain old Excel spreadsheet. My first
> experience doing this was modeling hydrostatic drive systems for
> Tunnel Boring machines for the English channel (Lotus 123 back
> then). I then moved to VF drives on the same application and then
> to Vector drives. Creating a spreadsheet that incorporates motor
> performance (varying swash plate angles at pump and motor, then # of
> poles, frequency, breakdown torque rating, voltage) and seeing how
> the motor reacts is immensely valuable. Plug in the various system
> inertias, gearbox ratios, efficiencies, etc. and now you can swap
> motor/drive technologies and play what if. After you have done a
> few dozen models for different applications / drive technologies . .
> . you get a much better understanding of the strengths and weakness
> of each.
> The fact of the matter is that very few people really understand all
> the interrelated systems involved in a motion control application.
> Trajectory planning, axis coordination, (current / torque / position
> / velocity) loops, system resonance, stiffness, backlash,
> synchronization with external events, etc. all are factors that come
> to bear on a motion / drives application. Some requirements are
> more stringent than others, but unless you really know the
> application, you will not know which ones are most important.
> It is best to have a strong theoretical background so that when
> confronted with "real" experience that you have a chance of
> correlating observed behavior with a reasonable explanation of the
> laws of physics governing that behavior. I have yet to attend a
> class or course that has done a good job of pulling it all together.
> The danger of presenting a class that supposedly "covers it all" is
> that you give people a false sense of confidence. The best college
> classes that I have to fall back on were the Newtonian physics
> classes and System dynamics classes. This has been followed up by
> much in the way of experience designing custom drives and
> integrating drive systems. IMO . . . the more you know, the lower
> your estimation should be of how much you really know. I learn
> something new about these things we call drives every day!
> Hope this helps,
> Ken Brown
> Applied Motion Systems, Inc.
> http://www.kinemation.com