[edclayton1]

Can anyone tell me if the accuracy of stepper motors is really adequate for building a decent quality plasma CNC table? I was reading on PlasmaCam's website that they chose to use servo motors for their "x" and "y" axis as the stepper motors can lose too many steps and without the encoders the problems compound themselves as the cutting continues. Can anyone enlighten me on the benefits of using hybrid steppers as well? As I see it right now I tend to think PlasmaCam is probably correct and should follow their lead, but this will cause my cost to go up more than I hoped. Any info would be greatly appreciated. Thanks

[plain ol Bill]

Blah, blah, blah - they all talk their yada yada yada. Steppers have been used on more tables than servos ever thought about with great results. With that said and out of the way both types are great. Can't help you with any hybrid info.

[datac]

Nothing wrong at all with a properly designed, setup and operated Stepper system... NOTHING ! Much of it is "servo hype".... Regards most machines built or used by people in this forum, "servo's" are often just a new "carrot" to dangle in front of people to get them to loosen up their wallets. MILLIONS of machines have been built with steppers that have performed flawlessly over their operational lifetime... many of them performing medical "life supportive" operations as well. Next time Grandma is attached to one of those little boxes dispensing meds at her bedside in the hospital, there is a larger percentage chance it is using a stepper than a servo !

Lost steps and position ? Well, it certainly could be said that a company could design a natively crappy control system that indeed could cause quality drivers to not interpret or misinterpret step and dir commands, but honestly, a GOOD control program simply will not do that. Likewise good quality drivers and good motors will also not fail. Doesn't say much for what the Plasma company is claiming about their software ???

Accuracy.... I believe you can find one of the hobbyists most informed step motor driver builders comments somewhere right here in this forum, pointing out that steppers are actually far, FAR more accurate than servos..... kind of why servos need the feedback from encoders to start with.

I'm not bashing servo. Servo certainly has its place and its equivalent successes. Some machine designs will demand servo simply when rpm limitations of steppers might not provide a high enough travel speed with enough torque to accomplish the intended work.

Generally, people who have completely failed with stepper and heralded their successes with Servo and then blamed steppers for their problems simply did not have their particular stepper system designed or set up correctly.

I USED to feel that steppers where always noisier than servo, but some of these newer stepper drivers are amazing.

Hybrid motors ? Partially new hype..... I mean, they are essentially just new and improved methods and manufacturing standards that have made the motors perform better... stronger, etc. Obviously one should buy the latest and proven greatest, but that label "hybrid" is pretty much just a fancy word as far as I'm concerned.

[bhurts]

I agree that steppers are just fine for a lot of setups. The very first Nc machines made in the 60s were stepper machines. That said stating that steppers are better or more accurate is crazy. It seems to me that if that were true all the truly industrial machinery would not use servos. Since they do I will fall in there camp that while servos are not needed for every application they are a better fit for high accuracy, speed, and reliability in industrial environments.

Ben

[Bubba]

Some time back, Mariss (Mr. Gecko) did an analysis of stepper vs servo and it basically boiled down to power required to make the system preform.
If the power requirement was ~<100 watts, use steppers. If ~>200 watts, use servos and if in between, buyers choice!

I won't dig into the formulas here, but there on the zone some place.

- - - Updated - - -

Some time back, Mariss (Mr. Gecko) did an analysis of stepper vs servo and it basically boiled down to power required to make the system preform.
If the power requirement was ~<100 watts, use steppers. If ~>200 watts, use servos and if in between, buyers choice!

I won't dig into the formulas here, but there on the zone some place.

- - - Updated - - -

Some time back, Mariss (Mr. Gecko) did an analysis of stepper vs servo and it basically boiled down to power required to make the system preform.
If the power requirement was ~<100 watts, use steppers. If ~>200 watts, use servos and if in between, buyers choice!

I won't dig into the formulas here, but there on the zone some place.

[plain ol Bill]

Here is a pic of a pc. I cut yesterday I 3/8 w/ my stepper system. It just don't get much better than this,

[datac]

>>>>That said stating that steppers are better or more accurate is crazy.

While I never said "better", but I did say "Accurate". And that statement is at least 99% true. Only continuous advancement in servo technologies keeps me from giving it 100%.

Accuracy Meaning: The degree to which the result of a measurement, calculation, or specification conforms to the correct value or a standard.

Servo's have an inherent hard time with 100% positional accuracy simply because they can not escape oscillation. Of course we all know that oscillation has continuously been minimized over the last 20 years. More on that in a minute....

It was Mariss that offered up an awesome explanation in one of the forums some years ago about this exact issue. I don't think he is "crazy". I think he is a brilliant fellow. His comments made me at the time, do even more reading because of the constant stepper verses servo arguments. My "Reading" was not in forums of hobby machine builders who will forever argue one over the other, but documentation from long time, unaffordable to us little guys, tried and true motion control manufacturers. It's certainly a deeper subject than many here care to delve into, and for the machines we typically deal with here, not really that necessary to delve into.

Nonetheless, I think the basics were that a high quality stepper motor, manufactured under strict methods have, and will always have an inherently "built in" absolute ACCURACY by nature of the design. If one controls just ONE thing.... that being input frequency, the step motor is rock steady, stable, and dead nuts "centered" in a magnetic phase when stopped. In motion, there is no "wander" or oscillation by design, and the most microstepping methods along with newer motor designs pretty much take care of the rest. They have no need for encoders to determine "accuracy" (not to be confused with the potential of using an encoder in closed loop operation simply for positional confirmation).

The above all falls apart if by chance one gets an imported "high quality" knockoff that is, well, anything but "high quality". PLENTY of that going on these days.


>>>> It seems to me that if that were true all the truly industrial machinery would not use servos. Since they do........


Ya,.... but no. "All - Industrial Machinery" is a really big paint brush..... I assume your mostly referring to the machining center world, and true... they do pretty much have that whole world running on servo systems for a lot of very good reasons, one of which is they can achieve their set goals with it. Impressive speed, impressive torque at speed, and "adequate" accuracy. Note that most "accuracy" claims are seldom challenged, or seldom measurable by the end user anyhow !

I would hope that we could at least agree that there can be a level of accuracy that is required or adequate for one thing, but a far different level necessary for something else. In other words, while a major machining center manufacturer might claim insane accuracy, it does not mean that some application might need an even higher accuracy... like the grinding of a space telescope lens for example. In the end, a machining center manufacturer shoots for an adequate target.... not what might be possible if money was no object.

Beyond that world of machining however, the Industrial, Scientific and Medical applications that demand extremely stable, highly ACCURATE positional control are still primarily stepper. And, there is nothing wrong with that. Video applications have primarily used stepper.. not so much because of accuracy but more regards stability.


>>>I will fall in there camp that while servos are not needed for every application they are a better fit for high accuracy, speed, and reliability in industrial environments.


The world of machining centers use servo today because it fits the bill. And no surprise, it fits the bill better than it had 20 years ago because they have been able to develop systems that provide them with what they are required to have regards accuracy rather than the idea that they are the most accurate possible. The majority of advancement is in faster and better electronics. The worst inherent issue with Servo is going to be oscillation at rest, but don't discard the fact that they also have oscillation in motion. Encoders are supposed to always monitor position and allow the control to "adjust" for where it thinks it should be. So, it's really only with faster and faster processing, thanks to better electronics and better software development, that oscillation has routinely been minimized. Its not gone.... but it has certainly been minimized into impressive numbers.

In the end, if the OP's concern was that PlasmaCam makes the claim that they use servo because steppers routinely fail for them.... well, they kind of show that they might not know HOW to set up a stepper driven system ? Not much for a warm fuzzy feeling if you ask me. I think plain ole Bill summed it up pretty well with "blah, blah, blah !

[xjdubber]

It really depends on how you use the technology as well, with a proper servo system using feedback to achieve coordinated motion, they will be more precise then a stepper is. However when using step and direction signals a servo is no more precise then a well setup stepper system as they simply follow the same signals. There is definitely a reason why most all industrial machines use servo's and that is because of coordinated motion where feedback from servo's is used to adjust position and speed at the same time thus being able to machine very accurately. However with plasma load is not really changing that much. I believe the statement before mentioning power requirement is a valid one. Torque and rotor inertia with speed will determine what type of system you should look at. Most importantly though if you want to achieve accurate motion, you need to minimize backlash, this will determine how precise your machine will be.

blah blah blah blah.....

- - - Updated - - -

It really depends on how you use the technology as well, with a proper servo system using feedback to achieve coordinated motion, they will be more precise then a stepper is. However when using step and direction signals a servo is no more precise then a well setup stepper system as they simply follow the same signals. There is definitely a reason why most all industrial machines use servo's and that is because of coordinated motion where feedback from servo's is used to adjust position and speed at the same time thus being able to machine very accurately. However with plasma load is not really changing that much. I believe the statement before mentioning power requirement is a valid one. Torque and rotor inertia with speed will determine what type of system you should look at. Most importantly though if you want to achieve accurate motion, you need to minimize backlash, this will determine how precise your machine will be.

blah blah blah blah.....

[bhurts]

I read the faq page at the geckodrive website. I'm sorry datec but I see nothing there that supports your point of view over my own.

Ben

[jimcolt]

Accuracy of a properly designed stepper system vs a properly designed servo system is essentially the same. The big difference between steppers and servos that has a noticeable effect to plasma cutting is with the torque vs speed curves between the two motor / drive designs.

A stepper has a relatively narrow RPM range with high torque....in other words you have to gear (belt or gearbox reduction) a stepper so that it spends most of its cutting time at a speed that is in the range of its best torque. Why is this important? Generally on thicker materials plasma runs at reasonably slow speeds.....however when you get into gauge thicknesses....plasma needs to run at high speeds (200 to 600 ipm for best results). When you are cutting parts with a lot of fine details and small holes....typically steppers have sluggish acceleration when asked to move fast (because of their torque curves)....and this sluggish acceleration keeps the speed slow on small, intricate details of the cut program. You can gear a stepper to operate well at high speed......but then usually you will get some oscillation and rough cuts when trying to run at low speeds. A "properly designed" (matching the gear ratio, the motor size and the weight of the moving parts) stepper machine usually performs acceptably throughout normal speed ranges.

Servos have a relatively flat torque vs RPM curve. They have good acceleration capability at low speeds and high speeds. The matching of the motor size to the moving mass is still important, but has less of a detrimental effect with servos as compared to steppers. Servos are more complicated with their feedback (encoder) loop, so they cost more. Because of the torque / speed advantage.....you will find servos on all high end industrial machines.....some of these large machines can have gantries that can easily weigh over a ton....steppers would not work well. Another advantage of servos is the ability to always be able to get back to an exact position....the cnc knows the location.

I am a servo fan, however I have seen hundreds of stepper machine designs with air plasma systems that cut just as well. I have both in my home shop.

Jim Colt Hypertherm

Can anyone tell me if the accuracy of stepper motors is really adequate for building a decent quality plasma CNC table? I was reading on PlasmaCam's website that they chose to use servo motors for their "x" and "y" axis as the stepper motors can lose too many steps and without the encoders the problems compound themselves as the cutting continues. Can anyone enlighten me on the benefits of using hybrid steppers as well? As I see it right now I tend to think PlasmaCam is probably correct and should follow their lead, but this will cause my cost to go up more than I hoped. Any info would be greatly appreciated. Thanks

[Torchhead]

We are agnostic in the "religion" of motors since we provide both types. Lots of improvements over the years have made some of the issues with steppers using quality drivers (low speed cogging. resonance, etc) go away. The size to power ratio has increase 2 or 3 fold in just the last 6 years (Hybrid steppers are part of the reason) . It then starts to boil down to Physics.

Simple comparison list

The RESOLUTION of a system does not predict the ACCURACY. Over a certain point other factors play a much larger role. I submit that any advantage of increased servo accuracy is negated by rolling bearings on square tube which cannot match the accuracy of a true rail system, but with plasma the difference between a final accuracy of .005 and ,001 is of little value. Servos get their higher resolution primarily from the fact of the final step down gearing

After you can hit max cut speeds (most cutting is done at 350 IPM or less) the most important number is acceleration. How FAST you can get from 0 to full cutting speed defines how well you can track the actual tool path. Systems with poor acceleration yield rounded corners and skewed detail. Since acceleration is defined as A = F/M (basically the linear force applied to the moving load divided by the weight of that load). The heaver the load the more force you need. Whether you measure acceleration in milliG's or IPS/sec units, there are minimums you have to hit to get cuts that follow the defined toolpath. Having .005 accuracy is not worth much if the tool can only track within .050. As cutting speeds go up so must acceleration. A table that can cut 3/8" flat steel at 30IPM may flake out cutting 18 ga that is badly warped.at 350 IPM.

We typically recommend servos for multi use tables (mixed contact and non-contact cutting) or heavy loads where the higher RPM of servos let you run higher gear rations and boost the rotary torque by factors of 10 or 20. That is why you see servos on bigger heavy industrial machines although you can still find Bridgeports from the 80's.with big honkin' steppers.

TOMcaudle
www.CandCNC.com