PWM DC brush motor control using PLC output?

P

Thread Starter

Patrick Allen

I'm currently looking for a way to control a variable voltage (9V-18V) DC brush motor. The spec calls for the motor to be PWM controlled with at a frequency of 200Hz +/-10% With a duty cycle ranging from 50% to 100%

Since we'll be using an SLC 500 (probably 5/03) for other controls anyway; I was looking for a way to incorporate it.

I had hoped for a Plug and Play module. But I can't seem to locate such an animal. My second thought was to use an analog output module to send a variable voltage to control the duty cycle.

One problem is that we need to control 8 of these motors on one fixture. I'm hoping to avoid designing and building "one off" boards in this case, as that sort of thing often comes back to haunt us later.

Any ideas or suggestions are greatly appreciated!
 
G

Gerald Moore

Do you already have a motor drive? Or is that one of the things that you are looking for?

Can the motor manufacturer offer a solution?
 
P

Pierre R. Hinse CET

go to www.vantec.com, they make PWM amplifiers for robot systems. If you ask for a modified RBSA, you will be able to drive the amplifier with an analogue output from your PLC. Ask to speak to Dale.

A satisfied Vantec customer,

Pierre Hinse
 
K

Kaufman, George

Since you have 8 motors to control from one PLC, maybe you should consider indexing servo drives with a DeviceNet interface. Check
www.MotionOnline.com for the indexing drives. A-B has DeviceNet master cards for the SLC 500. All 8 drives can be on one DeviceNet link and can be separately addressed.

Best regards,
George Kaufman
 
P

Patrick Allen

We have not yet chosen a motor drive. The greatest difficulty that I am having with that, is meeting the customer's spec. of a 200Hz PWM frequency. I imagine this to simulate the real world controller. (which,unfortunately I cannot get)

Each motor will run through a specific load profile. I think Magtrol has a good solution for that. Although, the techincian's idea of "running a touque transducer in reverse" as a controllable brake...I'm not sure if something like this is possible. But because the profile includes operating the motor in different directions, the Hysterisis Clutch set up we've used in the past likely wouldn't work well.

I'm still gathering details from the various engineers involved. But I really appreciate the feedback I've recieved on this board.

Patrick Allen.
 
M

Michael Griffin

Are you making a dynomometer system for testing automotive motors? This is what it sounds like from the motor specs. If this is the case, you may have to simulate the actual application closely to ensure that there are no "unknowns" introduced to the test process.

A typical automotive PWM system chops a DC voltage using what they call a "solid state relay", although the device is not normally just a
standard SSR. Both Motorola and Siemens (and likely others) make these devices for automotive applications. These things are being used more often these days instead of resistors to get multiple speeds. A PWM signal is produced by either the engine controller, or by some other means, and the "PWM device" (I don't like calling it an SSR) just follows it, amplifying the signal. The reason you can't get the "real world controller" may be because the actual PWM signal is generated by the engine controller.

However, if this is a test system, you may need more features than are present in a standard device. You may find that the current limiting or short circuit protection are not adequate from an off the shelf device (you are expecting to occasionally hook up bad motors). The Siemens device is actually sort of a big chip to which you can add various components(resistors, capacitors, etc.) to get optional characteristics. If you go this route though, you are basically talking about a custom circuit anyway.

I suspect that in the end, you are going to end up with either a custom power supply, or a custom PWM chopper fed from a lab power supply. A
few questions you might want to ask your product engineers are:
1) Is the 200 Hz +/- 10% mean they actually want to vary the PWM frequency by that much, or is that just the acceptable tolerance?
2) Is the test to be conducted (or repeated) over the full voltage range (9 - 18VDC), or is that just the possible voltage range in the automobile? In other words, what is the actual test voltage?

Depending upon just how complex this test becomes and how automated it is, you may find that controlling it with a PLC may be difficult and not suit your needs. A PC may give you more test interface possibilities (GPIB, data aquisition boards), and also the ability to log results. Testing 8 motors at once with one PC may not be very easy though, unless you are doing
exactly the same thing with all 8 motors at the same time.

As for your dynomometer brake, if you are interested in asking us any questions about that you should look into what the speed, torque, and
test duty cycle are.



**********************
Michael Griffin
London, Ont. Canada
[email protected]
**********************
 
P

Patrick Allen

Thanks for all your feedback Micheal

"Are you making a dynomometer system for testing automotive motors?"

Our customer seems to be undecided as to whether they want a performance tester, (ie. dyno) or a durability tester.

"Both Motorola and Siemens (and likely others) make these devices for automotive applications. These things are being used more often these days instead of resistors to get multiple speeds."

"You may find that the current limiting or short circuit protection are not adequate from an off the shelf device (you are expecting to occasionally hook up bad motors). The Siemens device is actually sort of a big chip to which you can add various components(resistors, capacitors, etc.) to get optional characteristics."

TI makes a chip (TPIC2101) that is pretty good controller in an application like this. The rest of the circuit would be relatively simple. However, I would still prefer a turn-key solution for this. It will make supporting the equipment much easier...especialy if it ends up in another country.

I put pretty much exactly the questions you mentioned here to my customers, and am waiting for a response. My initial contact indicates that they want to keep things as simple (cheap) as possible. The only real question right now, is whether they want to base cycle counts on time, or on hall effect counts. (We're thinking the latter will require high speed counters. 6 pulses per rev. approx. 5000rpm 4 motors per PLC) I'm starting to wonder if the SLC (our default choice) is going to be the best solution here.

Again, Thank you for all the feedback.

Patrick Allen
 
M

Michael Griffin

Patrick Allen wrote:
<clip>
>Our customer seems to be undecided as to whether they want a performance
>tester, (ie. dyno) or a durability tester.
<clip>

I guess this will make a big difference to what you do. A durability tester could probably be a lot different than a dyno. It also makes a big difference whether this is a lab (R&D or QC) tester, or a production tester. I assume this is a lab tester, because otherwise why would you be considering durability testing?

>TI makes a chip (TPIC2101) that is pretty good controller in an
>application like this. The rest of the circuit would be relatively
>simple. However, I would still prefer a turn-key solution for this. It
>will make supporting the equipment much easier...especialy if it ends up
>in another country.

I understand the desire for a standard off the shelf component, but I'm not aware of anything which does what you want.

Some possibilities and problems are:
1) Use a single power supply, and 8 conventional solid state relays which won't mind being switched at 200Hz. This doesn't give you individual current limiting or voltage regulation.

2) 8 individual custom regulated power supplies with a PWM modulated output. You have already said you don't like this.

3) 8 standard lab power supplies, each feeding a standard SSR. You would get individual current limiting, plus you would be able to adjust the voltage individually to account for resistance losses in the leads. I can't recommend an SSR which would be suitable. I'm also not sure whether the power supplies would like having their loads switched like this.

4) 8 standard lab power supplies which will let you enable/disable them at 200Hz to produce square wave outputs. This might be possible, but I can't give you a part number. I don't know what the minimum rise time would be, so I'm not sure if this would work even if the power supply were OK.
I have used 4 quadrant power supplies which were capable of being modulated at up to 20kHz, but these were sine wave outputs. These are also quite expensive.

It would be worth while contacting a power supply company to see if they can recommend something. Xantrex is pretty good (www.xantrex.com) and reasonably priced. You may be able to find suitable specs from their web
site. If you are not familiar with lab type power supplies, I think you will find them more expensive than you expected.


Some of the test criteria you may need to consider could be:

- Individual voltage regulation of each motor. You may need to apply a very specific voltage at the motor connector. You don't want resistance losses in the wiring to affect the applied voltage. If you are feeding several motors from a single power supply, you may not be able to find a set point which satisfies this criteria for all of them.
- Individual current limiting of each motor. You don't want the failure of one motor (overload) to affect the rest.
- Rise and fall time of the PWM modulated voltage. You want this to be limited by the motor inductance, not the power supply output response.

I don't know how important each one of the above is for your particular application.

>I put pretty much exactly the questions you mentioned here to my
>customers, and am waiting for a response. My initial contact indicates
>that they want to keep things as simple (cheap) as possible. The only
>real question right now, is whether they want to base cycle counts on
>time, or on hall effect counts. (We're thinking the latter will require
>high speed counters. 6 pulses per rev. approx. 5000rpm 4 motors per PLC)
>I'm starting to wonder if the SLC (our default choice) is going to be the
>best solution here.
<clip>


**********************
Michael Griffin
London, Ont. Canada
**********************
 
Top