Motion Controllers

All of the work I've done to date has involved writing custom code for custom boards -- and it works pretty well.

Now I've got a customer who may be best served with an off-the-shelf motion controller. My problem is that there are just too dang many of them! It seems like everyone and their brother make these things!

My technical needs are pretty basic -- an analog command input and an analog feedback need to be subtracted, with PID applied to the error signal and an output of either a PWM or analog drive to an amplifier (an on-board amplifier would be nice, but not necessary). Because the command and feedback are both velocities, the ability to monitor a position signal for limiting the drive would be nice but probably not necessary.

This would be for an OEM product, so just a board should be sufficient. If I'm not mistaken size won't be a problem unless the thing is implemented with 5881s (and they'd have trouble with 12V anyway). What _is_ important is that I have some flexibility in tuning it, the things are easy for my customer to configure once I find the right tuning, that I don't have to spend a bunch of time writing custom software for the thing, and that the product won't be obsoleted (or the company fold) before my customer is ready to move on to bigger and better things.

Anyone have any suggestions? I can google as fast as you can, what I'm really looking for are companies or products that you have direct experience with and can recommend with confidence.

Reply to
Tim Wescott
Loading thread data ...

On Tue, 08 Nov 2005 22:30:17 -0800, Tim Wescott proclaimed to the world:

Tim I did some work recently that involved an indexer by YASKAWA. While the company was not to great with tech help the rep was and the product was good. I don't have the reps name and address here but need to retrieve it at the job site anyway.

The servo and controller works in both velocity and position with all the feedback and amp built in. Pretty universal package. The only problem I see is the analog. My setup involved sending a command to go to a programmed position. The position was selected by BCD input not analog.

I feel the rep could answer your questions quick and dirty but I suspect he works regionally and you may be outside his territory. Someone else may have a better suggestion but I will get his name in the next day or so and post it to you if you want it.

What is the analog set point signals form and function. Is it a velocity or position signal? What type of motor do you want to control or are you looking for the servo too? You just didn't give enough info for me to nail it. Since you work mainly in custom board level controls you may not be familiar with some of the standard universal motion packages available today.

If you are looking for a simple PWM DC controller, Mineric makes a whole line of good OEM boards that take 1-5/10 vdc inputs that I have used a lot of with good results. If memory serves they go up to 3 hp on 90/180 volt motors. They are inexpensive and designed for the OEM. After thinking on what you wrote (and left out) I suspect that this is what you are looking for. They have armature feedback built in or tach for tighter control.

Want to give me a little better description of the app?

Be well,

HoP

The preceding message represents personal opinions and/or advice that may prove incorrect or harmful. But then maybe not. Feel free to disregard.

------- Words have no Warranty ------ ------- No View without Merit ------

Reply to
HoPpeR© trading at 1492¥

On Wed, 09 Nov 2005 03:11:35 -0500, HoPpeR© trading at 1492¥ proclaimed to the world:

Sorry the board manufacture's name is Minarik. Here is the link.

formatting link
Be well,

HoP

The preceding message represents personal opinions and/or advice that may prove incorrect or harmful. But then maybe not. Feel free to disregard.

------- Words have no Warranty ------ ------- No View without Merit ------

Reply to
HoPpeR© trading at 1492¥

Tim Wescott wrote in news:cO6dnWhYod3tCuzenZ2dnUVZ snipped-for-privacy@web-ster.com:

Tim-

formatting link
I've generally been pretty happy about NI products. The question w/ their stuff will be how much of an off the shelf solution can they offer you, and how much of their software will you be tied to. That said, the product won't obsolete--NI has always been good with backwards compatiblity for their hardware. There's a pretty big user base, and lot's of sources of help when you need it. Also, the system is likely to meet all of your customer's needs for some time to come-- very flexible stuff.

The downer is the expense--there will likely be a computer in your loop somewhere.

The great thing about NI is their sales force. Contact your local rep with your specs, and see what he can wrap up for you. Your time commitment to look down this path will be minimal.

Reply to
Scott Seidman

Hmmm, analog input, analog feedback, analog output, simple PID. Sounds like an opamp to me. Servo amplifiers with tach loops have had these capabilities for the past 30 years. What advangages are gained by putting this in software? That said, most digital amplifiers such as AMC's digiflex series have these capabilities built in. Are there any resonances that might require notch or lowpass? Changing system parameters that would require periodic retuning? Are monitering/reporting functions required?

Reply to
jeff

I can throw out future expandability and do it with a couple of op-amps (and a 555 if I need PWM). In fact that's one of the options I have in mind. The customer sorta gasped when I gave him the price to develop, however, and I believe that he doesn't have in-house board manufacturing capability, so I'm looking for something I can just buy.

I wouldn't say no to an analog controller (servo board?).

Reply to
Tim Wescott

Thanks Scott. The thought of an NI-type solution running on PC-104 boards did cross my mind -- it's absurd overkill for what the customer needs at the moment, but it does put a PC in their product right where they'll need it in a few years.

Reply to
Tim Wescott

It's angular velocity for both feedback and setpoint (command input, really, from a joystick attached to a user). It's to add gyro feedback to a pan-tilt table which has DC motors but no other closed-loop control. It doesn't need to engage in any autonomous motion. It just needs to turn as fast as the joystick tells it to, maintain a gyro-aligned pointing angle when its base moves, and not burn up the motors when it runs into a stop.

No kidding. About all I know is that they're out there, they're well accepted, and there's a gazillion of them. In fact, when I originally started talking prices with the poor guy I had assumed a custom solution so implicitly that I just quoted a development price without thinking that this could be done with an OEM board. I could almost hear the "thud" of him hitting the floor in his return email, and I'm glad that he didn't just drop the conversation right there.

It is this lack of familiarity that leads to the post -- I want to get something that will do what the manufacturer says it will and won't disappear tomorrow.

I'm thinking I also lack some terminology as well -- would the thingie that implements the PID loop and possibly an amplifier be called a "motion controller" or a "servo"? In custom systems the thingie that issues commands is either called the "executive controller" or the "#$%@!"; the thingie that implements the PID is called the "compensator", and it's usually followed either by an amplifier or a string of curses from a more experienced engineer. Nowadays the usual way to implement all this is in software, with an appropriate DAC and ADC and an amplifier tacked on out there in the real world.

I've described most if it above. The only tidbit that I've left out (that I know of) is that it'll be a small production run OEM thing; having a board that could be easily configured in a repeatable manner would be ideal. I don't really care whether the configuration involves setting switches or downloading parameters on RS-232 or dancing around the thing widdershins just as long as it is easy to do in a light industrial setting by someone who isn't an electronics expert.

I think a simple PWM DC controller (servo?) will be just the ticket -- I'm going to look at your link as soon as I hit the "send" button.

Knowing what I know about his market I suspect that as soon as the customer implements this he's going to get requirements from _his_ customers that make him wish he had a custom solution, and I'm going to (gently) tell him so. None the less starting out with an off-the-shelf solution may be a good choice for him at this time -- he can prove out all of the other system additions, gain street cred, then move to a custom or semi-custom controller at a later date.

Reply to
Tim Wescott

Tim Wescott wrote in news:tuOdnRWoKN4euu_eRVn- snipped-for-privacy@web-ster.com:

If you end up following this road, pay attention to the TXI platform. I've been very happy with it.

Reply to
Scott Seidman

Scott Seidman wrote in news:Xns9709884E6F1DCscottseidmanmindspri@130.133.1.4:

Ooops, PXI, not TXI

Reply to
Scott Seidman

On Wed, 09 Nov 2005 09:11:38 -0800, Tim Wescott proclaimed to the world:

From what I see now Minarik is just the ticket. The boards are under $200 buck and the ones I used 10 years ago are still available. I would go ahead and use the PWM regen unit because it has all the capabilities and options built in that they might decide to use in the future.

One note. Some of the board's inputs are not fully isolated and this can cause problems sometimes if you are using several boards. I just used a isolator block with the ones I used that did not have it built in.

Be well,

HoP

The preceding message represents personal opinions and/or advice that may prove incorrect or harmful. But then maybe not. Feel free to disregard.

------- Words have no Warranty ------ ------- No View without Merit ------

Reply to
HoPpeR© trading at 1492¥

Tim, my company makes motion controllers that can do what you want. This is a good candidate

formatting link
Peter Nachtwey Delta Computer Systems, Inc.
formatting link

Reply to
Peter Nachtwey

I looked at the link -- they should have the amplifier I need, but I'm still without a controller. Overall, I need to implement the following loop:

Plant /-- Controller --\ Amplifier (DC motor) _ .-------. .------. .-------. 0-5V / \ | | | | | | ------>| + |--->|Hpid(s)|---->| ka |---->| k/s |---o---->

\_/ | | | | | | | ^ '-------' '------' '-------' | | | | Gyro | | .-------. | | 0-5V | | | '----------------------------| kg s |

Reply to
Tim Wescott

First I decided to change my Nic, but this is the poster formerly know as HoPpeR. Several reasons which I will not go into unless someone wants to know. I would have done this from the beginning but was not for sure if I was going to stick around. Sorry for the inconvenience.

I'm having problems with the ASCII drawing for some reason I try fixed font and get the same. Got any suggestions? You might just send me the thing via email in html or pdf.

From your description I do not understand the need for the controller. The amp has a controller and the boards I used accept voltage inputs. The board I was thinking about takes a scaled or unscaled speed control signal and produces a proportional motor speed. The PID and voltage or encoder feedback is built in. It should be a complete control/amp package. I looked at Peter's company and his stuff does the same thing. His product line is more high end and I suspect more expensive but will do the same thing. If they truly have up to speed tech support 24/7 and the price is not too much higher, I suspect that I will be buying my boards from him. He seems to have evolved in the logging industry. Peter, if you are listening in maybe some input would be good here.

If you are trying to control the gyro application is something like synchronous application you might need to think about some sort of more sophisticated controls, but I don't think this is the case if the input is by a joystick directly. I used these boards on a web system with about 10 motors all sharing the pull load. The product was thin rubber sheeting and the operators trimmed the speed by had while the master speed was simply a single voltage source. I think that this is an example of one or both us use misunderstand some fundamental things because of different terms like I posted about earlier.

Reply to
A. Paul Montgomery

If it seems that someone is being dense just assume that it's me. I looked for indications of PID compensators and didn't find anything. Could you give me some specific part numbers as examples, or do all of their drives include PID compensation?

I had my head up my assumptions in another way that may prove troublesome, however. In production at least the system has to run off of 12 or 24VDC -- in hindsight it's obvious that I should have mentioned this. This doesn't mean that I can't use AC powered stuff to help the guy out for a prototype/demo system, however (assuming I can limit the voltage to the motors, of course).

Reply to
Tim Wescott

Hi Tim,

You have not said what kind of motor you are trying to drive (brush/brushless/induction/stepper/SR/other), what the feedback is (tachometer/encoder/resolver/ferraris/other), nor power levels (W/KW/MW/GW) nor the operating voltage for this device (Battery/50/60Hz AC/400Hz AC/Single or Three phase/other). I did not spend the time going through the Miniarik site, but expect that they do have suitable products. Virtually ANY modern servo drive can set up a velocity mode pid loop internally, either as a discrete implementation, or as a digital loop. I have been using some drives from AMC that are fully programmable digital drives that can operate in torque, velocity, or positioning mode. Take a look at

formatting link

Reply to
jeff

Reply to
mr converter

Yarg. Yes, I am leaving out a bunch of detail -- it drives me up the wall when other posters do that, why do I do the same thing? Some is scattered in other posts, but:

  • DC brushed motors.
  • Gyro (rate) feedback.
  • 10's of watts, possibly 100 or so.
  • 12V or 24V supply rails.

As I mentioned before you can give me a pile of parts and some FR-4 and I'll make it do exactly what you want, it's just knowing what to buy off the shelf that's overwhelming me with choices.

Reply to
Tim Wescott

They all do.

This is the big limiting factor here. Industrial controls are all going to be 90 or 180 vdc with the supply to the drive controls at 120/240. Drives for low voltage DC are going to be a different class all together. I'm sure they are made but this is something I never work with.

Reply to
A. Paul Montgomery

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.