Coarse/fast screw pitches with sevor motors?

[gearsoup]

Correct me if I'm wrong, but fast pitch lead/ball screws have always been discouraged for use with stepper motors. I believe it's because of the 1.6 deg movement of the stepper motors forces you to have a rather large movement per step.
But.....

Servo motors use encoders. Doesn't an encoder have a very large counter that gives you better than a 1.6 deg resolution? Motion Control Products - DMM Technology is what I have in mind.
If they do have a large resolution, then would it not be feasible to run, say, a 25mm screw with a 25mm pitch? My machine is going to be roughly 5ft square, and I really don't want to be limited to low rapid speeds. I understand that screws are somewhat limited in their maximum rps, due to whipping, so perhaps a coarse pitch is where the answer lies. Any thoughts?

To add to the discussion, my machine will be a CNC router designed for wood only. I have no intention of machining metal, other than light engraving work. Any serious cutting will be done in wood. I also want to be able to do some very detailed 3d cuts.

[ger21]

Correct me if I'm wrong, but fast pitch lead/ball screws have always been discouraged for use with stepper motors. I believe it's because of the 1.6 deg movement of the stepper motors forces you to have a rather large movement per step.

I've been here 8 years, and I've never seen their use discouraged. They're just not used very often. Over the years, the trend has been towards higher lead screws.

Back then, 1/2-10 acme was common, but has progressed to today when 1/2-10 5 start is the standard (.5", or 12.7mm lead) with some people using 1/2-8-8 start, with a 1" pitch.

The majority of rack and pinion users get about 1" of travel per motor revolution, equivalent to a 1" pitch screw. But rack and pinion is more efficient, transferring more power.

Anytime you use steppers, there's a tradeoff. The faster you want to go, the less resolution you'll have. This is due to the fact that stepper motors usually have a maximum usable rpm of 1000 or less. You can't just spin the motor faster to move faster, you have to move farther with each motor revolution, which reduces resolution.

Servos, on the other hand, have a few advantages when used at higher speeds. As was mentioned, they have higher count encoders, which can give you 5 times or even more steps per revolution than you get with steppers. And since servos tend to have much higher rpm's than steppers, they are typically geared down, which provides more torque, and even more resolution. The downside to this, is that you may need to provide much higher step rates to a servo, sometimes requiring additional hardware to provide these faster step rates.

Generally, when using servos and ballscrews, you don't drive them direct. You want to gear the motor, so that at your max feedrate, the motor is spinning at 80% of it's rated no load rpm.

[gearsoup]

Hmmmm....gearing down. Now that's not something I've even considered into my design. There is room for it, just need to get creative. That would be small pulley on the motor and big one on the ball screw, correct? or is it the other way around...

[ger21]

Yes, the motor spins faster than the screw.

[mike_Kilroy]

......Servo motors use encoders. ............then would it not be feasible to run, say, a 25mm screw with a 25mm pitch? .........

You seem have a good handle on steppers and their use, so I will ignore those comments. You ask a simple question above; there is NOT a simple answer to it. the positioning you can do with servos is down to a few counts of the encoder feeback, so yes, if you have 1 enc count per 1 um travel, you can tell the servo to go there. but this is not the issue. the issue is inertia.

when you dont have the mechanical advantage of lots of turns per inch, the reflected load inertia back to the motor gets big. Exceed about 5x the motor inertia and you will have a nice mechanical system that you likely cannot control well enough to satisfy you. at 25mm/rev pitch, you will for sure have to be cognizant of this inertia mismatch issue. at 10 turns/in the load inertia is reduced so much reflected to the motor that anyone can put a servo on an axis and make it work.

screw whip is a function of length, bearing support method (single or dual each end) and rpm. so yes, large pitch can get you more IPM w/o getting into pitch problem. but the inertia issue still has to be addressed.

The majority of rack and pinion users get about 1" of travel per motor revolution, equivalent to a 1" pitch screw. But rack and pinion is more efficient, transferring more power.

not sure where this came from; no, rack/pinion are no more efficient than ballscrews. a ballscrew should be 95% eff: a 3" pinion will move about 10" per rev so you need a 10:1 gearbox on it to get down to the stated 1"/rev - the rack/pin will also be about 95% eff, the gearbox at best will be 95% so the total is .95*.95= 90% total - hardly "more efficient" than a ballscrew!

And power transfer is not an issue in servo or stepper systems. In fact, if you want the math, max power transfer happens in a system where the load inertia exactly matches the motor inertia - power transfer has no meaning to us trying to figure out what motor to use unless we are running off solar cells and batteries in the middle of Zaire or something.

gearing was mentioned. Gearing may be very important. You can either get the load inertia reduced to <5x the motor inertia by picked the proper screw pitch or if necessary for other reasons (like cant get max IPM w/o whip unless use 1 or 2"/rev screw for example, or you have a free 25mm/rev screw) or by judiciously adding the proper gearing between the motor and screw in this case. direct drive to the screw is ALWAYS best if possible to reduce the extra backlash and compliance of a gearbox or timing belt. Only add gearing to a servo system if necessary to solve some issue.

ok, all this said, I hope you can see just throwing a high pitch screw and some random gearing and servo to an axis is likely to not make you a happy camper at all. The problem is you will not be able to adjust servo bandwidth to anything that will hold the axis in position against machining forces - you may have experienced these kind of axes - you push on the table and it oscillates back and forth a dozen times before stopping.

BUT - you CAN figure what to do using the hardware you may have available, including that high pitch screw! and it only costs you about 1 hour of your time!

Download the free Kollmorgen Motioneering software.

http://www.kollmorgen.com/en-us/supp.../motioneering/

with it you can enter all your parameters like ball screw pitch, diam, length, gearing choices, and let it pick a servo motor size for you. Once picked, you are free to go to gecko or whoever you want with the data and get the correct motor. but by playing what-ifs and changing the gear ratio, screw pitch etc etc etc you will see how each effects your axis. you will then be able to make an educated choice of hardware that has a good chance of working really really well! Of course as a Kollmorgen rep we would be happy to quote various motors and drives to get r done too!

[louieatienza]

gearsoup, I'm using 1/2"-8, 8 start on all three axes on my machine. Sure my "resolution" is only .005" but for most of everything I do in wood it's good enough for me, and better than what I could achieve by hand! I've even been able to do light aluminum work as well. I don't think I'd be making parts for NASA, but, I've always been amazed by the elvel of detail I could achieve despite the resolution...

As mike_kilroy mentionrd, there are issues when the cutting forces exceed the holding torque of the drive motors. I keep it in mind and run I'd estimate to about 80-90 percent of their limit. I've been able to cut pretty aggresively without ever stalling my steppers.

So it depends also on exactly what and how you want to cut. As far as rapids, I think the one advantage the r&p drives have is no whipping issues.

[gearsoup]

Material being cut is wood. I figure 99% of the machines use will be in wood. 2d outlines, 3d surfaces, lots of engraving for signs, etc.
Because my machine is 4ft square, I'd really like to see something over 400 ipm for rapids. That may seem greedy, and it very well could be. At 393 IPM (figured on using a 10mm pitch screw) it would take me almost 8 seconds to get from one edge to the other. Not shabby. But on the other hand, that is with the motor at max speed (DMM tech motors) Most likely I'll be down to the 300 IPM or so. Meh. Can't afford to buy screws twice. Probably just end up going with the 10mm pitch screws and living with a somewhat slow machine. After I make some cash on it I'll upgrade to better screws, might even get some new-old-stock ground screws.

[louieatienza]

Material being cut is wood. I figure 99% of the machines use will be in wood. 2d outlines, 3d surfaces, lots of engraving for signs, etc.
Because my machine is 4ft square, I'd really like to see something over 400 ipm for rapids. That may seem greedy, and it very well could be. At 393 IPM (figured on using a 10mm pitch screw) it would take me almost 8 seconds to get from one edge to the other. Not shabby. But on the other hand, that is with the motor at max speed (DMM tech motors) Most likely I'll be down to the 300 IPM or so. Meh. Can't afford to buy screws twice. Probably just end up going with the 10mm pitch screws and living with a somewhat slow machine. After I make some cash on it I'll upgrade to better screws, might even get some new-old-stock ground screws.

Gearsoup, it may not be as good as direct-drive, but have you thought of using their smaller servos and gear them down? Then you should get pretty good power, and speed as well as resolution...

[gearsoup]

Let's think about this for a minute. Not being sarcastic or anything, just trying to set things right in my head. If you use a smaller servo, gear it down for increased torque, you do that at the cost of speed, no? All I can compare the whole idea to is a 10 speed bike with a derailer (spelling sucks) great torque, but low speed, good for climbing a hill.
On the other end of things, you could gear up, but sacrifice torque.

Am I looking at things back-asswards?

[lgalla]

The smaller servos,400Watt are 3,000rpm.With your 25/25 ball screw you would get 3,000ipm.Gearing down will give tons of torque with decent speed.
Larry

[gearsoup]

Hmmm, and conversely, using 10mm pitch screws, I could gear 2:1, doubling my torque, but still giving me 580 max IPM. Hmmmmmm. that's not too shabby. Need to run some numbers and see if those motors can even support my machine.

[mike_Kilroy]

........At 393 IPM (figured on using a 10mm pitch screw).............But on the other hand, that is with the motor at max speed (DMM tech motors) Most likely I'll be down to the 300 IPM or so. Meh. ..........

probably just end up going with the 10mm pitch screws and living with a somewhat slow machine. ......

I'm curious why you back off on top speed prediction from 393 to 300 by saying the 393 is at max motor speed? is it a problem to run at max motor speed?

You had mentioned screw whip earlier; you want 400ipm but expect 300 because you cant go full rated speed on the motor(?) what is the whip speed? is that not a problem after all then with the 10mm/rev screw? do you have a chart that shows whip speed per dia/length? and are you aware this speed goes up significantly with dual bearings at both ends? Lastly, do you need/want another quote for reference on screws if you don't already have them? as a thomson distributor I would be happy to pop u a quote for reference to show what is out there.