495,938 active members
8,968 visitors online
495,938 active members

# Thread: Ideal stepper motor rpm?

1. ## Ideal stepper motor rpm?

Possibly a dumb question, but, with regard to the torque curve shown in http://www.cncrouterparts.com/cdata/...rque_curve.pdf it may seem that the best rpm to run that stepper is; well, as low an rpm as possible.

However, based on the fact that horsepower is proportional to rpm x torque, I calculated the following (rpm and torque values from the spec sheet above):

Code:
```RPM	Torque(N*m)	power (no unit specified)
60	5		300
120	4.6		552
240	4		960
360	3.6		1296
480	3		1440
600	2.4		1440
720	2		1440
840	1.6		1344
960	1.4		1344
1080	1.2		1296
1200	1		1200```
Based on that, the ideal range for this motor looks to be between 360 and 1080 rpm; indeed if I actually wanted a 60rpm output, it would be better to run the stepper at 600rpm, and gear it down by 10:1. Given (unrealistic) 100% efficient gearing, the 2.4Nm at 600rpm would become 24Nm at 60rpm.

Am I on the right lines here? My question is probably "in general" in relation to steppers rather than this particular model; I have no prior experience with them, but I'm considering having a go at a DIY CNC build.

2. ## Re: Ideal stepper motor rpm?

You'll find that you need the most torque during acceleration. Since acceleration is linear, you need the same amount of torque at your max speed as you do when you start from a stop. This limits you to certain speed, where you have enough torque to accelerate at a given rate.
Some people accelerate faster, and limit their top speed. Others accelerate lower, so they can run a bit faster.

3. ## Re: Ideal stepper motor rpm?

Originally Posted by ger21
You'll find that you need the most torque during acceleration. Since acceleration is linear, you need the same amount of torque at your max speed as you do when you start from a stop. This limits you to certain speed, where you have enough torque to accelerate at a given rate.
Some people accelerate faster, and limit their top speed. Others accelerate lower, so they can run a bit faster.
So if I can gear a stepper motor, and get (what I believe to be) sufficient torque even at the highest output rpm I need, then I should be OK at other (lower) speeds?

My idea is related to a CNC lathe (whereby moderately large objects would be turned, and cut with a spindle). Realistically the sort of rpm figures I'd need for the lathe axis would be between 0.5 and 10rpm (maybe 20rpm at an absolutely maximum). A 60:1 reduction ratio would give some pretty impressive output torque figures (even allowing for efficiency losses in the gearing). Or would it be better to go for a lower reduction ratio, and not push the stepper up to ~1000rpm (given that I'd have sufficient output rpm and torque even with a 30:1 ratio)? I guess the question there is what rpm range is usually best for stepper motor life?

4. ## Re: Ideal stepper motor rpm?

Stepper motors should last "almost forever", as the only wear parts are the bearings, and steppers spin relatively slowly.
You really just need to do the math to see what works best for your application.

5. ## Re: Ideal stepper motor rpm?

Originally Posted by ger21
Stepper motors should last "almost forever", as the only wear parts are the bearings, and steppers spin relatively slowly.
You really just need to do the math to see what works best for your application.
Thanks Gerry. The question is more in my ignorance of steppers (I've only ever used machines with servos). If you can get the required output torque and speed (with appropriate gearing) and keep the stepper speed to, say, 500rpm max, is that better than gearing it to use the full "rev range" of the stepper? I keep hearing about issues of pushing steppers to higher rpms, so I'm trying to understand if 0-500rpm is preferable to 0-1000rpm (assuming that 0-500rpm is considered adequate for the application).

6. ## Re: Ideal stepper motor rpm?

For a given machine velocity, the faster you can spin the stepper the better.
Since steppers have a fixed resolution, spinning twice as fast effectively doubles the resolution, and can result in smoother motion, especially at low speeds.

7. ## Re: Ideal stepper motor rpm?

Originally Posted by ger21
For a given machine velocity, the faster you can spin the stepper the better.
Since steppers have a fixed resolution, spinning twice as fast effectively doubles the resolution, and can result in smoother motion, especially at low speeds.
Good point - I hadn't thought of that. Well; should I ever get round to building this machine then I don't think I'll be pushing any limit on steppers; it doesn't need big speed or significant accuracy (0.2mm / 8 thou would likely be fine). Many thanks.

8. ## Re: Ideal stepper motor rpm?

Once it's running, you'll always wish it were faster....

9. ## Re: Ideal stepper motor rpm?

Originally Posted by ger21
Once it's running, you'll always wish it were faster....
Ironically I might need to make it slower!

My 3-axis CNC machine can cut MDF (1/4" dia cutter, 1/4" deep slot) at around 2 inches per second (120ipm). Getting that sort of linear feed speed on a large turned object (e.g. something 12"-16" in diameter) requires a very slow rpm. A ratio of 36:1 on the servo I'm looking at will just about get me a good speed range for anything from 2" to 16" diameter, but as I'm mostly planning on diameters of 8" and upwards I may need to slow it further. This also assumes I could make the machine stiff enough to cut at 120ipm (so it may need to be slower still).

Anyway; it's not for a commercial project (so long cutting times aren't a problem), and with the large reduction ratios it appears that I should have plenty of output torque in order to not stall the stepper whilst cutting (at least, based on what little understanding I have of figures in that area).

EDIT: Oh, and I can get acceptable resolution with 1/4 stepping, even on a 16" diameter object; which I assume should be no problem from the point of view of the control electronics

10. ## Re: Ideal stepper motor rpm?

Hi,
stepper motors are a class of 'reluctance motors' and are best considered constant power output devices. The torque reduces (very) approximately
linearly with rpm for a (very) approximate constant power output.

I bought (second hand, EBay) 23 size Vexta 5phase steppers with integral low backlash (<2 arc min) 10:1 planetary gear reductions.

Firstly five phase steppers exhibit increased resolution (500 full steps per rev) compared to two phase steppers (200 full steps per rev). They maintain their output torque
somewhat better than two phase steppers. I run mine at 2400 rpm with good authority, ie no loss of steps whereas many two phase steppers at that speed have lost that
much torque that any load is likely to cause lost steps.

Five phase steppers are less inclined to vibration and resonance than two phase steppers. As you know most people run two phase steppers run a microstepping
regime of 8, 10 or 16 micro steps per full step in order to avoid midband resonance. I run my five phase steppers WITHOUT microstepping and have NO vibration
or midband resonance issues.

The gear reduction means I get great output torque (705 oz.in) from these small motors. Of course the gear reduction means that the output shaft is only spinning
at 240 rpm with the motor running at my self imposed speed limit of 2400 rpm. When directly coupled to my 20mm diameter 5mm pitch ballscrew it means my
rapids are only 1200mm/min or 48 IPM. The thrust of the axis is very high, over 1500lb force at stall. I have my acceleration set at 375mm/sec2.
I can set it quite a bit higher but my machine tends to thump backwards and forwards on the bench at higher accelerations. I had it tied to the wall at one point!

The gear reduction means I get great thrust (cutting force + acceleration) and acceleration but at slow speeds. My machine is designed to mill steels so slow
is not unreasonable but I would like to go faster.

I strongly support Gerrys comment that high acceleration is very important to cycle times, as much or more so than high max speeds. Additionally when in CV
mode, the norm, highest possible acceleration maintains the closest toolpath following accuracy.

As good as these five phase steppers have been they are a reasonably rare and expensive beasts. There are only about four or five manufacturers of five phase steppers
around the world, all quality brands and none of them are cheap. The low backlash planetary gearboxes (made by Apex) are likewise very expensive.
Despite buying second hand they cost about \$300 an axis (stepper including gearbox and 230VAC Line input 5 phase drivers). I have had great use from them
but I paid a high price for it.

If I were designing and building my mill today I doubt I would select the same units......for just a little more I could have 400W DMM AC servos for instance.

You are correct, if you use gear reduction you can achieve very good results with steppers at the expense of speed. Its not the end of the world....I use my machine
a lot despite being fairly slow. Unfortunately low backlash gearboxes are not cheap.......a good quality name brand like Apex could be several times the cost of
the stepper and drive. In short if you can find some good/reasonable gearboxes at a modest price your idea could work well. I suspect that once you do some
research you'll find that the cost of the gearbox is prohibitive.

Craig

11. ## Re: Ideal stepper motor rpm?

Very useful info - many thanks.

From what I've seen of torque vs RPM graphs, the torque remains fairly constant; up until some fairly low RPM, then it drops away - roughly linearly with the RPM (i.e. maintaining constant power). With a higher voltage input, that flat torque "shelf" is wider. I can guess at reasons for that, but my electrical/electronic knowledge is limited.

From what I've been calculating, I could get adequate resolution and torque for my application with 1/4 steps and a 36:1 reduction ratio. I assume there's no downside to configuring a stepper controller to use finer steps - so long as it can be driven at an adequate pulse rate?

Gearboxes are a concern. I've just finished reading through this (https://www.cnczone.com/forums/linea...ary-table.html) massive thread. It seems to me that GT3 belts might actually be the most sensible and cost effective solution - given that 0.1mm (4 thou) accuracy would be fine; unless of course any backlash would lead to significant drift in true position.

The intended application is effectively turning large objects (that will have some non-symmetric sections - hence not doing it by hand on a lathe). I'd hope then that (CAM-permitting) there wouldn't be that many rapids and repositioning of the cutter; so conservative acceleration shouldn't be a problem.