1. ## Servos and things

I am in the process of building a 24" x 30" router basically for fun, since I have a commercial unit already. I have utilized linear rails and ballscrews from a previous commercial machine I had, so this thing will be stout and light on friction. That being said I want this thing to scream as far as rapids and cut speeds. The ball screws are an aggresive .5" lead so here in lies my questions. Much has been said about steppers but I am going to use servos.
How do you gauge the maximum amount of rpm you practically use on a servo motor even at rapids? I have read that you never will reach the maximum and don't want to, so is there a percentage that is the comfort zone or calculating zone for gearing?
I intend on using geckos, so my next question is power supply. I know for steppers you take the rated voltage and increase exponentially, how about with servos?
There are a wide spectrum of motors available and keeping in mind rapid moves, what criteria is felt is best? Smaller motors with higher rpm geared appropriately? Larger slower motors less ratio on gearing or even direct drive?
I currently have in my possession 36v 2.3 amp 3000 rpm motors, and 7500rpm 9.0 amp 24 volt motors which is better?
My next question is in regard to encoders. I haven't really done any research but I will ask here anyway. Do you base the count on a relationship to motor speed or do you want it to tie in to actual travel (did that make sense)?. If I use a higher speed motor with a higher gear ratio the motor will of course turn more revolutions to move the router a set distance than would a slower motor with a lower ratio.
Any other thoughts on setting up a servo system would be appreciated, Mike.

2. Hi Mike

You have asked a lot of questions in one go

Volts times amps = watts, which is a measure of how much work the motor can do, so:
36v * 2.3amp = 82.8 watts
24v * 9amp = 216 watts

So your 24 volt motors are capable of producing the most power, if you have a power supply that can hand it over.

But, your coarse pitch screw and high rpm motors are kind of working against one another, because I doubt you want to move at 3750" per min in Rapid This means a fairly substantial gearing down to achieve the proper range of output. But, that is not necessarily a bad thing. If you use timing belts for reduction, you can always change the ratio after you have the machine built.

You need to check the maximum encoder counts your controller will be able to process, as this may influence where you mount the encoders, as well as how many lines of resolution they would need.
An encoder mounted directly on a high rpm motor will generate a lot of pulses, and may overload your controller's feedback capacity, limiting your upper rapid speed unexpectedly.

However, mounting your encoder directly on the screws is considered more accurate. This will allow a higher resolution encoder to be used. Higher res encoders cost more, of course.

Whichever way you mount it, the servo PID settings can be adjusted to accomodate the system. Servos operate according to the amount of error between the present position and the commanded position ( with an upper limit of course). The output voltage is proportional to this error, so the controller really could care less what the servo/ballscrew gear ratio, or what the encoder ratio is. When the encoder feedback shows that the machine is getting near the commanded position, the output voltage is cut back smoothly and quickly to zero, as the commanded position is reached. The PID parameters adjust the way the servo motor settles into position, controlling overshoot, undershoot and oscillation.

What this boils down to is the accel/decel ramps that you will set up are going to be the ultimate limitation for your chosen drive amps. Acceleration is typically when the controller commands the amp to output the maximum voltage. A sharp current spike here will trip out the overload protection on your drives. The actual inertia and friction of your machine slides produces the limitation, and I don't know exactly how you would know in advance what you will be able to achieve in real life rapid speeds.

3. Thanks! I didn't ask these questions, but your tutorial has brought a clearer understanding to me too!

4. Thanks Hu, you maybe see my lack of clear direction. Although 3200 inches a minute is intriguing, I am thinking in the 350 to 400 ipm range on rapids. So with that in mind a 10:1 or similar timing belt reduction should give a smaller motor but fast motor good torque. The part that I am still unclear of is how much of that 7500 rpm will I practiclly be able to use. If I am shooting for a set ipm rate, what of that 7500 rpm do I really have or do I want? The encoder mounted to the screw makes sense and might make for an easier utilization of some of the motors I have. Mike

• Mike,

I think you might be comparing stepper motor performance to servo, that is, steppers tend to crap out and get weak when they are really going fast.

Servos do not have this limitation. They will run the nameplate rpm if you can supply the voltage. You will not be harming the motor by doing so.

The one consideration is the actual ramp up time to get the motor from a standstill to 7500 rpm, but this is usually not a major problem

• Power is RPM times torque. Sero motors develop their rated power at about 80% of their no-load speed. That would be 6,000RPM for you.

Figure out what maximum RPM you want your leadscrew to turn at. Divide that number into 6,000RPM. The answer will be your reduction ratio.

Mariss

• Mike,

You must also consider the step rate-of-delivery. Are you going to control these motors with a PC and using Step and Direction software? If you are, then the bottle necked may be the PC and the step rate that the PC and software can send. Depending on the encoder count and the step rate delivered, this my be your limit for max IPM.

• Mariss thats the magic number I think I wanted. I have a Flashcut Cnc signal generator currently running a mini mill, that will provide the step and direction. It is pc based and I will upgrade it to their new lightning chip. I don't believe this should be a problem. Mike.

• just thought i'd write a short reply to make sure that you know that those ballscrews can't take 6000 rpm. There is a maximum roatational/load formula you need to apply to the ballscrews to keep them from literally burning up.
if you are a commercial boy, just grab you mcmaster/grainger and look it up.

The easiest way to do rapid moves with servos is to use a timing belt driven by the motor through a pulley, one idler pulley on the other side, and the carraige bolted to the belt. I have built machines that do this, one was a reciprocating slicer, one was a laser cutter.

Just remember that speed costs accuracy, how fast do you want to go?