HELP!Direct Drive? Stepper/Servo?

[diyengineer]

http://www.cnczone.com/forums/showthread.php?t=100925
Above is my current Design.

I need help deciding if servos or steppers would work better.
I plan on using a 2.2KW spindle.
I only plan on doing aluminum and other non ferrous metals.

I have read you can directly connect the stepper to the ballscrew via a coupling and no gear reduction but you can't do the same with a servo. I like the idea of a closed loop system so its accurate, but i can live with steppers too. In my above link if you check out some of the recent renderings i am using Nema 34 size steppers/servos directly connected to the ballscrews. If i wanted the most power out of either stepper or servo should i plan on using some kind of gear reduction system to boost torque? I like doing things correct the first and since i plan on doing mostly metals, more accuracy the better along with power. Please someone give me a general break down of the pros and cons of using a stepper or servo on my specific machine, taking in to account i want to machine only metal (aluminum sheet .063" and 1/4 plate).

Thank you for your time, hopefully someone will be able to shed some monster light on the subject for me!
I have been looking at all the stepper/servos from www.kelinginc.net

[RICHARD ZASTROW]

As usual, it comes down to $$$$$$$$. I have to qualify this, I deal with higher powered machines.

My preference is permanent magnet brush less servos. Full torque at "0" rpm. When possible, hollow shaft motors such as Danaher Direct Drive Cartridge motors or Bosch/Rexroth frame less motors. These can be attached directly to ball screws or spindle shafts without external couplings. BUT they are pricey!!!

You have to consider the total cost of the axis drive. Mounting brackets, coupling, belts, pulleys etc.

Generally, you will have less alignment problems, less complexity and a more compact assembly.

Dick Z

[ger21]

You need to figure out how much weight your moving, how fast you want to move it, and how much force is required. Once you know the required force, you can calculate the needed torque, and rpm to achieve your goals. If you go to the Gecko forum here, there's a FAQ that tells you how to decide what you need.

[Al_The_Man]

There seems to be alot of misleading information when it comes to servo V steppers.
Historically major CNC MTB's have always used servo's, with many in direct drive to the ball screw application, from a Mazak lathe Z axis moving a 250lb live tooling tool post, to a Weidemann Punch press whipping a material sheet around at several hundred inches/min

YouTube- wiedemann turret punch

YouTube- wiedemann turret punch

The machine in the video uses an AC servo direct drive on to a high lead ball screw.
The torque curves for both servo and stepper start at maximum at zero rpm, the servo tends to stay fairly flat along the total rpm range, steppers drop fairly quickly with increase in rpm.
I think a lot of the misconception arises due to steppers as a rule having a higher torque value for a given NEMA package.
One reason why servo's are often used with reduction is because they offer higher maximum rpm rates, reduction allows the use of an economically smaller motor while maintaining the required feed rate due to this higher max rpm.
Al.

[diyengineer]

so since i want to direct drive a stepper/servo using no reduction (and i wish i could afford AC servos but DC only) what would you use if you were me on a budget?

[stevespo]

I'll just reiterate what everyone else said. This is a complicated equation. You've got to balance your budget with whatever goals/needs you've chosen for your project, as well as what the physical limitations of the components are.

It's very worthwhile doing the calculations to ballpark the type of torque you will need to move your machine at the speeds (and cutting forces) you want to achieve. This is a nice tutorial, and there are other resources as well, including the Techno H835 catalog, etc.

What size motor? tutorial

Once you have a number in mind, or a range (for me, it was 60 oz-in to 100 oz-in, and 250 IPM), you need to see what can provide that number reliably at the speeds you want to move your materials (or machine) at. This is tricky because the torque curves available for stepper motors are usually provided with specific drivers, power supplies, micro-stepping, etc. So, it won't be exact, but may at least get you pretty close.

You may also need to factor in the complexity of the installation. Direct drive, verses timing belts, verses geared reduction, etc. The smaller DC servos (NEMA 23) may not develop enough continuous torque to meet your needs, unless you reduce them to gain some mechanical advantage. The bigger servos can be direct driven, but they may be overkill for your particular application. BLDC and AC servo systems tend to be a bit more expensive and may blow your budget.

You also have to make sure that your components are a good match for the speeds you're after. Finer pitch screws will give you higher resolution and more mech adv but will require higher RPMs to make speed. At some point you'll hit a critical speed and the screw will start to resonate (whip).

So, I guess my advice is to decide pretty firmly what type of performance you want to achieve. How fast and how much force is required. Then you can start to work towards how to get there. Motors, drives, power supply, screws, etc. Is the closed feedback loop important to you, or can you make due with a properly sized stepper setup? That is another very important question.

The alternate approach is to find a machine similar to yours and find out what they used to get their performance. Then copy it. They have already effectively worked out the details, and you stand a pretty good chance of success by using their design, provided your parameters overlap well enough.

With your specific machine, I don't know. You're not moving a heavy gantry around, just a spindle and a table. I wouldn't think your torque requirements would be huge, but it's hard to say without running the numbers. With ballscrews and low friction linear rails, you've got very good efficiency, so I think it comes down to "How fast do you want/need this thing to go?" and "Do you want/need feedback?"

Steve

[stevespo]

The other thing I'll mention is that your cutting speeds will probably be determined by your spindle, your tooling and the materials you'll be cutting.

To cut most materials effectively, you will have to maintain a "reasonable" chipload for the tool and material. A high speed spindle, particularly for soft metals is going to need fairly fast feedrates to do this.

How fast is that? What types of tools will you be using? How deep are the cuts?

Onsrud Aluminum Chart

For instance, running a .125" 2 flute endmill in .063" AL, at 12K RPM at a .003" chip load would require 72 IPM. That's a concrete number to start with, provided the machine is rigid enough to handle that load and the spindle can power though that cut.

If you want to run the spindle at higher speeds, and take a more aggressive chip load, you might be up in the neighborhood of 192 IPM. Look up the Datron machines if you want to see some very high speed machining in soft metals. 60K spindles and cutting speeds of 400+ IPM.

So, it really depends on what you want to do with it. If you want the most flexibility you'll have to design for a broad speed range to give you options.

Steve

[diyengineer]

Originally Posted by stevespo The other thing I'll mention is that your cutting speeds will probably be determined by your spindle, your tooling and the materials you'll be cutting. To cut most materials effectively, you will have to maintain a "reasonable" chipload for the tool and material. A high speed spindle, particularly for soft metals is going to need fairly fast feedrates to do this. How fast is that? What types of tools will you be using? How deep are the cuts? Onsrud Aluminum Chart For instance, running a .125" 2 flute endmill in .063" AL, at 12K RPM at a .003" chip load would require 72 IPM. That's a concrete number to start with, provided the machine is rigid enough to handle that load and the spindle can power though that cut. If you want to run the spindle at higher speeds, and take a more aggressive chip load, you might be up in the neighborhood of 192 IPM. Look up the Datron machines if you want to see some very high speed machining in soft metals. 60K spindles and cutting speeds of 400+ IPM. So, it really depends on what you want to do with it. If you want the most flexibility you'll have to design for a broad speed range to give you options. Steve

Thank you for taking the time to write all that Steve! Very insightful. I have been told that the Kelingin ballscrews are a .5mm per turn (16005 have a gear ratio.5mm per turn is a 5 to1 ratio.1000RPM on the motor is 250ipm on the screws with a gain in torque) want feedback, along with DC nema 34's keling sells most likely hooked to Gecko drives.

Specs below are to the servos ive been looking at, $139 Each.
http://www.kelinginc.net/KL34-180-90.pdf
They fully tell all the specs so hopefully some one can tell me if those plus the ballscrews and my ballpark weights will work and maybe a ballpark of how many ipm. Most of the 2.2KW-3KW spindles ive seen go from 6000-24,000rpm.
I would like to use the below drivers linked to the above servos as well.
http://www.geckodrive.com/product.aspx?c=2&i=14457

The y axis table itself (depending on what thickness i use) will weigh 30-50 pounds.

The X axis (which has to move the Z axis weight as well) will weigh i'm sure another 35 pounds (spindle, rails, bearings, plates, etc)

The z axis has the weight of the carriage plus the spindle and the motor so i assume 25 pounds or so.