DC variable-speed motor control 1

[albarrett]

Hello,
I'm in the process of designing a machining device that will utilize a 1/2HP DC motor (90 VDC, 1750 rpm) to rotate a machining surface. The speed of rotation will need to be accurately varied with the use of a foot pedal, in order to permit hands-free operation of the machine. However, it is important that the speed of the rotating surface be accurately controlled by the foot pedal, and not adveresly influenced by varying load or rotational inertia - constant speed under load and controlled acceleration/deceleration are essential. Are there specific types of PWM drives suitable for this type of application? Any assistance would be greatly appreciated. Thanks.

 

[cbarn24050]

Hi, a common thyristor dc motor controller will sufice.

 

[albarrett]

Hi, thanks for the reply. Though more economical, would the thyristor controller be better suited for this application than a PWM? For reversible rotation and controlled deceleration (under load with rotational momentum), would the contoller need to be regenerative?

 

[wired1]

I think the term "thyristor" is used synonymously with "electronic" in this area. Your application sounds well within the abilities of a modern electronic drive. All the major players have an offering. PWM is probably the
most common method of control. Yes, if you need controlled decel, you need regenerating. I would recommend contacting a drive distibutor app. engr. for assistance.

 

[cbarn24050]

Hi, do you really need 4 quadrant? 4 quad will mean a tachogen as well as a drive. You can still use a thyristor unit, it has 2 bridges.

 

[albarrett]

Thanks, wired1 and cbarn24050, for your feedback. I must confess, cbarn24050, that I'm not sure about 4 quadrant control. Adding a tachogen in addition to the drive seems excessive for this application, I agree.

Strangely, the electrical aspects of this project seem more complex than I expected. Having a a mechanical engineering background, I did not anticipate variable-speed control of a DC motor to be so elaborate. Getting a rotating object to speed up or slow down in direct linear proportion to pedal movement (regardless of varying load) has been quite interesting to say the least!

 

[jbartos]

Suggestion: Since there is a need to reverse a direction of rotation, two-quadrant controller is mandatory. Also, there appears to be necessary to have a properly designed the feedback loop to control the speed tightly without the inertia causing speed oscillations.

 

[wired1]

The complexity is a function of the performance you want from it. You want reversing abilities, so that's two quadrant control. You want controlled decel, so that's 4quad control. If you really don't need a controlled decel, then your controller could be simpler. But it seems to me that the choice of control will have a direct effect of the quality of parts coming off your machine tool.

 

[albarrett]

Thanks. Indeed, the quality of the machining will be directly affected by the precision and performance capabilities of the machine. Both reversing and controlled deceleration would be required in some machining instances, so 4-quadrant control appears to be mandatory. And as pointed out by jbartos, so is an appropriate control feedback loop to monitor speed tightly.

Part of the impetus to design this machine has come from past experience with similar devices not being able to accurately control the deceleration of the rotating machining surface when subjected to the momentum of the load. Accelerating and maintaining speed under varying load conditions has been relatively easy to control. However, when the operator adjusted the foot-pedal to slow down the work piece while being machined, a lag time always existed during which the operator would have to wait until the rotation slowed down, on account of frictional forces, to the desired speed. And stopping the workpiece presented a similar problem, as larger workpiece loads translated into greater rotational momentum and longer spin-down cycles. The current machine must be able to prevent this uncontrolled deceleration altogether, and under varying load conditions. Pedal position (operator input) should specify the required speed/accel/decel of rotation, with the controller and closed-loop sytem ensuring that these requirements are accurately met.

At the same time, of course we must consider cost limitations. And it seems that 4-quadrant control in a closed-loop system is the more costly option in this case. And, perhaps, the only option. In retrospect, it may have been practical to consider a variable AC control system.

 

[RadarRay]

If you don't mind I have some questions for you on your application. The above posters have given you enough information for you to make a decision but some of your statements are interesting.

Before I go on, I should say 20 years ago I designed the drives on a couple of roll grinders. These machines had to resurface the rolls for paper machines. The rolls being 10foot wide and up to 40 inches in diameter, therefore they were very high inertia machines.

While you state you must have precise control, I just can't see using a foot petal being that precise. Can an operator really set the speed and not vary it during operation.

With a motor so small as 1/2HP, how is it geared into the driven roll. And what are the typical inertia's you are driving.

What hasn't been mentioned, is if you use a DC motor with armature voltage feedback for control, you will typically not be able to set the speed regulator gain too high, especially on a thyristor drive. If the Interia of the system is large, you will want to make the gain high for good response. These two conditions do not interact favorably. It would seem to me that speed droop when loading the cutting tool against the load, could be one of the issues you have with the machine and armature voltage fb would surely give you more droop than a tach fb system.

So in some ways your application looks as though it is demanding good response and yet in others, it does not. At least based on the info you have given.

Just food for thought.