Stepper or Servo Motor?

Probably helpful to add in a rough approximation of torque needed.

Understanding the load kinematics is the most important thing to predict which approach will or even can achieve the needed results. Having said that there are several rules of thumb to consider.

A stepping motor requires a fairly complex bit of driver circuitry for it to work at all. A servo motor requires only a properly sized switch to run at its fastest speed. Getting a servo motor to move at variable speeds requires similar circuitry to what a stepping motor requires but usually with fewer outputs. Getting a stepper motor to move at a known speed or to a certain location is relatively easy without feedback. That is providing no stalling (aka missed steps) happens along the way that one runs the stepping motor. A change in the mechanical load on a servo motor without feedback can easily change the speed or final location. With feedback a change in mechanical load is easily compensated for with a servo motor. (Naturally with limits handled by feedback theories.) A stepping motor with feedback can also compensate for load changes. However this can quickly become a far more complicated problem with stepper size and encoder resolution concerns.

The answer was OK, but the complex electronics can be bought in quite small chips nowadays, they work out all the complex stuff for you. The L298D is a good cheap example of a stepper motor driver in a small chip.

For small servos there is the Mitsubishi M51660L chip for example.

If a really smooth operation is needed, the stepper motor should be selected for one with the smallest steps. that makes it more expensive generally speaking, as does greater size.

It can even then be "micro-stepped" with most modern controller chips.

Also, it can also be geared down to help further smoothing, which will increase the available torque as well.....

Though good power supplies will be needed with either solution, quality SMPSs are probably a must, efficient and reliable. Also cheap to buy on say ebay....

If you need something cheap and beefy, there are Youtube videos showing how a car alternator can be made into a big stepper motor quite easily....

See here:-

Converting an Alternator into a large Stepper Motor

I hope this helps.

One last thing, with motor cases of the same size the servo motor will usually produce more mechanical power.

This might work....note the pwm controller listed

http://www.amazon.com/Nextrox-120RPM-Powerful-Torque-Replacement/dp/B00B1KZ8UU/ref=pd_sim_60_4?ie=UTF8&dpID=41X8Z8IQ42L&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=0J51CQY5BBHVZ8F9PM8A

You have to tell us a lot more
Is this for a toy?
A servo does not go "around and around and around"
Why not use a 3-pole motor and gearbox.

A servo CAN go round and round, if designed properly for the job needed. Don't get fixed on Radio Control servos.....they are only part of the market.

Servo motors were used in large machines like big printers, to move the paper rapidly and very accurately for example.

Some servo motors are also called PCB motors as the main moving part is a PCB! This is to reduce the weight and increase the acceleration of the rotor.....

Its a big field that I have simply been out of for too long now.....

But if I did not need the accuracy, the power and the price of a servo, I would go for the stepper, also very accurate, but generally not as powerful for the same size....

As others have said, lot more information is required.

In my experience, steppers tend to run more smoothly than servos at low speed in direct drive.

The most common steppers I've used have 200 steps per revolution, but any good stepper driver can "microstep", moving partial steps instead of complete steps. Microstepping effectively multiplies the number of possible positions per rotation. The simpler microstepping drives can multiply by 2, 4, or 8, while more expensive ones can multiply by at least 64. Microstepping significantly reduces the 'jerkiness' of the motion, and reduces the noise level significantly.

In the few machines I've built using servos, I found tuning the PID factors a pain, but that was several years ago. Presumably they've improved automatic tuning since then. This is the main reason I lean toward steppers.

But the real question is what does your application need? Do you need to know the exact speed? Do you need to keep 2 or more motors synchronized? Do you need to know the exact position of rotation at any given time? Do you need a constant speed within the given range, or do you need a variable speed? How much is cost a factor? Is this for a device that may involve safety of the device or of the operator? Etc...

If your answer to most of those questions is no, then I'd suggest a brushless DC motor with a built-in gear reduction, similar to the one linked in an earlier comment. (I didn't pay enough attention in my brief look at the link to notice whether that was a brushed or brushless motor.)

He just wants it to go around and around and around.
He doesn't want any positioning or stopping and starting.
Just a 3 pole motor and gearbox is needed.
But he hasn't said how big the motor needs to be and how often it will be used or the supply voltage, so there's nothing we can advise.

OK, OK! This motor will go round and round. I have used small servo motors on projects that just go round and round. In fact, some of the better "drones" on the market today use servo motors, others may use cheaper motors but mini-servos are quite remarkable and there are chips that can drive them directly.

Obviously, I don't want to give up the farm for some initial considerations. The load experienced will vary from task to task. Run time per task may average 120 seconds. These things are not really important right now. The problem is just that there are so many possible solutions and I don't want to have to try them all.

Cost is, as always, the key. The bumpy ride of a stepper motor is OK if the steps are fairly small. Highest Torque is preferable, but at low RPM that starts to get complicated. It sounds to me like the lease expensive controls probably dictate some form of stepper motor. I don't know if steppers are made with lots of poles where several are in parallel for higher torque or not. That could be an answer but the demand would be very small simply because everybody wants a smooth torque.

How about this? Can I assume that in terms of the controller:

A) Steppers are usually driven in a digital fashion by H-Drive circuits per phase. With only 2 phases used in most steppers, are there designs that use sets of parallel poles for more evenly distributed torque about the rotor?

B) Servo controllers probably work better with more expensive analog (or PWM) drives. Is this a valid assumption?

C) Since a brush-less configuration is preferred, I assume the stator will require permanent magnets. For low RPM it sounds like it will require a lot of pieces. And I'm worried that too many pieces will just weaken the available field or require less clearance between rotor and stator. If this is correct, I have another option which I'm not at liberty to discuss at this time.

The rotary servo's I've seen are really powerful for their size. I think they will become very hot at low RPM. Stepper Motors can also be powerful but I'm concerned about losing and recovering sync under load.

I plan to use a separate encoder for feedback anyway because I want to detect slip (or insufficient grip). We are really talking about a design that is not currently available as far as I can tell. And I realize a public forum is probably not the best place to disclose every idea. I have rarely come up with any idea that has not already been invented or patented, but I just don't want to make it too easy to copy. Motor technology is probably hard to patent anymore because most ideas have already been documented. Maybe that is where I should look?!? I'm just looking for some good rules of thumb when it comes to the biggest torque for the buck at low RPM.

Sadly, you demonstrate many misunderstandings here about such systems.

One of the positive aspects of steppers is that they are brushless, if that is high on your list of priorities. The servos I worked on (years ago!) were not!! Maybe that has changed. VFDs are brushless.

Aligning or realigning steppers is easy and some people even add some form of feedback to tell them exactly where it is..I personally have never had the need myself.

As you have not I believe supplied us with any proper valid infos, we are clutching at straws....we are trying to help, but you are preventing us giving you valid help!! We still have little ideas what accuracy, what power and and and and.....

Please tell us everything you know!

I don't know why you marked your post as off-topic. It's not!

The ceiling fans are indeed a good example of the external rotor type of motor; I hadn't thought of them.

I just last week repaired a 30-year old motor that has roughly 200 contacts in its commutator, and roughly 100 poles. It was made specifically for very slow speed. Unfortunately, our current project requires a very steady motor speed of just over 1/2 RPM (just under 2 minutes per rev), and, although the motor can rotate that slowly, at that slow speed the torque is so low that it does not move steadily, so we had to add additional gear reduction. Here's a view of part of the rotor and commutator:

This is hardly a consumer product! I suspect a new one would be well into 5 figures, if it could be found. Since it predates the internet, like your experience, I have found only indirect references to it.

If you haven't already, look into "Printed Circuit" motors.

For, say a bicycle-rim sized motor, it would be perfectly possible to use only a few small permanent magnets between a pair of stamped pole pieces with on the order of 1000 or more poles as the external rotor. Hall-effect sensors (or equivalent) would be required to synchronize the fields with the internal coils. The difficulty for a real bicycle would be the high probability of distortion in the shape of the wheel. Very small gaps between rotor and stator poles are an absolute necessity for an efficient motor.

You mentioned you wanted a motor with a fixed rotor and moving outside. There are stepper motors with hollow shafts, that might address whatever your need is there, but I've never seen one where the outside could spin. That, of course does not mean they don't exist!

Although cost is always the key for consumer products, it is not always so for industrial and scientific tasks. I gather yours will eventually be a consumer product.

Generally speaking, High torque stepper motors simply have more layers of core than lower torque models, and so are physically longer, for a given motor size. They usually have two windings, with or without center taps, so either 4 or 6 leads. The circuitry for a 6-lead stepper driver is somewhat simpler than that for a 4-lead stepper, since the center tap is common, and always connected, but at any given instant, only half the copper in a given winding is being used, so a 4-lead stepper will presumably have higher torque. Since the complexity is no longer an issue with the incredibly tiny chips now available for small steppers, then the 4-lead type is preferable. The number of steps per revolution is determined by the number of notches stamped into the core layers; higher step numbers do NOT require more parts, just more little notches between poles. The permanent magnet is normally in the rotor, so no brushes are required.

Always, close tolerances between rotor and stator are required for maximum torque. This is true for all kinds of magnetically powered motors, including steppers and servos.

Since you've now stated that you already plan on a separate encoder, and want maximum torque at low speeds, then I believe your best bet is a DC Brushless motor with a built-in speed reduction gear set.

Like Andy, my peak knowledge on this was a few years ago, so investigate the newest stuff available, to contradict any of the above that may no longer be applicable.

You said:-

..... but I've never seen one where the outside could spin. That, of course does not mean they don't exist!

Have you ever seen an electric bike, look at one where the motor is in the axle of a wheel.....that is the exact motor he may need if he wants a slow VFD.

Its also 3 phase and brushless. Some have hall effect sensors for position information as well....some bike electronics need that, some doesn't!!

But as he has told us so little up to now, its like pulling teeth getting infos....

When I said that, I was discussing stepper motors. I don't recall having seen a stepper in external rotor configuration.

I have seen, and in fact owned, at least two external rotor motors. They were reel torque motors in a 3-motor reel-to reel tape recorder, where the rotors acted as flywheels to help smooth the motion.

In fact, I was thinking of an in-hub wheel motor as one of the possibilities that the OP might be trying to accomplish, but I discarded that idea due to his 1/2-2 RPS range. of course it's not at all clear whether he really means zero to one or the other of those values, or if 1/2 RPS is the minimum and 2 RPS is the Maximum.

I answered your point in our previous posts, about the availability of such motors, as you appeared to not know of them, but they are everywhere and very common nowadays..

....and that that the OPs requirements could easily be achieved by such a modified bike hub motor. VFD control is very flexible with such requirements. Bikes are "slow" machines. Furthermore, gearboxes can be used to up the torque and to allow a higher motor RPM to be reduced if needed.

Steppers are available today in many different configurations, certainly it may not be cheap in the configuration required, but if that is the only acceptable method, so be it......

As I have already mentioned, we cannot help the OP further until he gives out more detail.

I had the pleasure of fixing a problem with a machine that was tearing itself apart by replacing a motor/brake system with a servo/stepper motor with a shell that rotated whilst the rotor was fastened rigidly on a stand. It was a really interesting design in that had 512,766 (or so) unique stopping positions in one rotation. It was amazingly powerful as a 9" by 9" cylinder. Unfortunately, it cost on the order of $5000 with about $3000 for the controller. And it had some side loading issues that I had to be careful about but that was for a special problem.

For my application, my main focus is avoiding the need for gears. What I find interesting is that there are so few low RPM motors that can deliver much in the way of torque. I believe the problem is primarily a fundamental problem of motion and that the physics of the problem really requires an unconventional approach. With that being said, I'm leaning more the concept of a stepper motor with an unconventionally high number of poles. But I thank you for your contributions as I just needed to flesh this out a bit and try to avoid wasting a lot of time on undesirable concepts.

A direct-drive record turntable is about in this speed range, but torque and power haven't yet been discussed enough.