Need Help! Stepper Optimal RPM Range?

# Thread: Stepper Optimal RPM Range?

1. ## Stepper Optimal RPM Range?

Hi Everyone!

I have a basic question about what RPM range I should be aiming to use from a stepper to drive a ballscrew via pulleys.

If we look at a torque curve example attached at 24V, we see in the <400RPM range the highest and roughly constant torque. Then >400RPM we see a downward slope develop where torque drops rather linear with RPM increase.

Remembering that Power = Torque x RPM, to me this linear slope suggest maximum constant power has been developed and so torque drops as RPM increases.

If we directly drive a ball screw from the motor I think it makes sense to use the highest torque range.

However we have a pulley ratio in this system, that will convert RPM to torque, so I think I should be aiming to utilize the motors maximum power range, not maximum torque range. I should gear my system to use the >400rpm range.

Is my logic correct?

2. ## Re: Stepper Optimal RPM Range?

I thought to check my logic I would look at a Tesla power vs torque curve, because it shows both aspects. I think it confirms my thinking that maximum power is where the torque curve starts to drop off.

3. ## Re: Stepper Optimal RPM Range?

Also, it seems to me that a motor which reaches peak power at the lowest RPM is desirable because it will take the least time to get to that point. Example: a motor that hits 50 watts output power at 100 RPM will reach that power sooner than a motor which produces 50 watts at 500 RPM. Acceleration of the machine will be better.

I'm just looking for confirmation (or not) of my thinking here.

4. ## Re: Stepper Optimal RPM Range?

Hi,
comparing a Tesla motor to a stepper is incorrect. The Tesla motor is a synchronous motor whereas a stepper
is a variable reluctance motor. The physics of torque generation is different.

A synchronous motor will have near constant torque to its rated speed and therefater drops in the so called 'field weakening' mode.
A stepper has its maximum torque at zero speed and essentially decreases thereafter.

The principle determinant as to how badly the torque degrades is the winding inductance, the lower the better.
Many steppers have good or even great torque but also have commensurately high inductance, so while they have great torque
up to a hundred rpm or so they weaken badly at several hundred rpm. Manufacturers know that first time buyers look only at the
holding torque figure and don't understand inductance. As a consequence they build steppers with great torque figures but are hopless
at speed just to catch those first time buyers.

In 23/24 size steppers look for 1-2mH, 1mH preferred and reject anything over 2mH.
In 34 size steppers look for 2-4mH, 2mH preferred and reject anything over 4mH.

Often times, for a motor of a given size, there will be different models, some with high torque and some with lower torque. The lower
torque low inductance models are better for CNC and are often the high current versions of that motor size.

The classic way to overcome inductance is to use the highest possible voltage drivers you can get your hands on. At the current time
80VDC is about the practical maximum. A linear (transformer/rectifier/capacitor) power supply is preferred over cheaper switchmode supplies,
they have way WAY better overload characteristics.

This stepper has only 1.18mH and with an 80VDC drive could be expected to retain 45% of its rated torque at 1000rpm:

Object reference not set to an instance of an object.

Matched with this driver:

Object reference not set to an instance of an object.

With this power supply:

https://www.antekinc.com/ps-10n80-10...-power-supply/

You characterisation of a stepper as a 'constant power device' is reasonable. Consider however that a stepper is supplied with rated current
at all times. Thus even when it it not turning, ie producing no mechanical output, it is still absorbing supply current. How does that fit with
your characterisation? Your constant power characterisation is useful in the sense that it describes the torque/speed curve but is dubious
for calculation purposes.

Craig

5. Originally Posted by joeavaerage
Hi,
comparing a Tesla motor to a stepper is incorrect. The Tesla motor is a synchronous motor...
Craig

I was asking quite generally without a specific design mentioned as I thought if I understand the logic I can apply it to my situation. But I'll give my specific circumstance.

I'm making a very sturdy and well damped 3D printer. It doesnt have a huge amount of weight to shift but Inam hoping for rapid acceleration to some pretty fast speeds, like 500mm/s (40,0000mm/min).

I haven't calculated the mass or built it yet, but let's say the gantry and head weigh 500g not including ball screw inertia. It has 20mm pitch ball screws. It will be driven by 45V 1.5A rms drivers.

I know that's not a lot of current or voltage. But it is a really nice closed loop servo (uStepper S).

I plan to use Nema 23 but obviously my current and voltage is limited.

(On a side note, I would be happy to use closed loop stoppers from leadshine or others, but I worry about the quality of the drives. I know the drive signal purity directly reflects in the surface finish of the print. Trinamic driver ICs have a really clean output (stealth chip/ silent step) whereas many CNC router drivers are rather crude by comparison you can hear the drive signal in the motor coils. But maybe they have got better since 10 years ago when I built my router?)

6. ## Re: Stepper Optimal RPM Range?

Hi,
some of your target performance figures are very ambitious and I think are beyond steppers performance envelope.

For instance you say you want to achieve 40m/min and you are using 20mmpitch ballscrews. That would require the ballscrew
rotate at 2000 rpm. With direct coupled steppers thatis 2000rpm....which is achievable....just......but not with a 45V power supply.

I use Vexta 5-phase steppers with genuine Vexta drivers that take 230VAC input for a DC link voltage of about 150VDC. That allows my steppers to spin
at 2400rpm, but they have very little torque, that could not be direct coupled to a ballscrew. In my case the steppers go through a 10:1 planetary (low lash)
gearbox.

Achieving 2000rpm with a stepper AND still have useful torque is right at the very outside of stepper performance. If you use a gear or belt reduction
such that the stepper is only required to do 1000rpm you might have more luck, but the stepper will require double the torque.....back in the too-hard

Manufacturers of closed loop steppers would have you believe that they are faster, more powerful and never lose steps....total BS. Closed loop
steppers do have a couple of advantages (full step interpolation and realtime following error monitoring) but are bought at considerable price.

From your description of your project you need AC servos. Delta (Taiwanese manufactured in China) and DMM (Canadian manufactuered in China)
are two good quality brands that won't break the bank.

A 400W Delta B2 series (160,000 count per rev encoder) has 1.27NM torque to its rated speed of 3000rpm, and has a ten second overload of 4Nm.
In field weakening mode it will spin to 5000rpm. It'll EAT any stepper ever made. I got one of these and three 750W B2 series including one with a brake
and they are superb. They really make steppers 'so last century'.

Craig

7. ## Re: Stepper Optimal RPM Range?

Hi,
you may see in some other threads some calculations that predict the acceleration that a given torque can produce:

https://www.cnczone.com/forums/servo...ml#post2372136

If you follow the derivation and importanly the calculation that applies to my new build mill you will see that despite a very heavy axis (115kg) the acceleration
is determined NOT by the axis mass but the rotational inertia of the ballscrews. As it turns out the mass of the axis is usually a small proportion
of the acceleration potential....its more about the rotating components.

Lets make a few assumptions and guesses and do some calculation that might approximate your machine.

I suggest we take the D57CM31 Leadshine stepper I linked to previously. The spec sheet reveals a first moment of inertia:

J= 0.84 X10-4 kg.m2

Lets assume also that you have an 80VDC driver and supply such that the stepper can retain 50% of its rated torque at 1000rpm, ie 1.5Nm.

Lets guess you use a 16mm diameter ballscrew of 20mm pitch and 400mm long.

Lets guess that you make the axis as light as possible, say 2kg.

The first moment of inertia of the ballscrew is:
mass= (0.016/2)2 x PI x 0.4 x 8000(kg/m3)
=0.64kg
J= 0.64 (0.016/2)2/2
=0.205 x 10-4 kg.m2

The first moment (equivalent) of the linear axis is:
J= 2 (0.02/2*PI)2
=0.203 x 10-4 kg.m2

So the total first moment of inertia is:
Jtotal= (0.81 + 0.205 +0.2) x10-4 kg.m2
=1.21 x 10-4 kg.m2

So you can see that the rotational inertia of the motor and the ballscrew dominate the acceleration equation with the mass of the axis
representing less than 20%

Continuing with the calculation:
Angular acceleration = torque/ first moment of inertia
=1.5 / 1.42 x 10-4

The linear acceleration of the axis:
A= (10563 / 2.PI) x 0.02
=33.6 m/s2

Thus the stepper I've linked to could easily accelerate your axis but at 1000 rpm the torque is down to 1.5Nm even with an
80VDC driver and supply. As a matter of practice I doubt you could go much faster without losing steps.

Craig

8. Originally Posted by joeavaerage
Hi,
you may see in some other threads some calculations that predict the acceleration that a given torque can produce:

Craig
Again, thanks so much for your reply and walking me though the math.

I will give it some thinking over.

I had not realised rotational inertia was such a big factor! Kinda wishing I orders 12mm dia. ball screws instead of 16mm now!

There are a number of products offered by Delta and DMM, how much money are we talking for a basic 1.5nM drive and motor?

9. ## Re: Stepper Optimal RPM Range?

I dont think I can afford AC servos and I also dont think I can easily incorporate them with standard 3D printer control boards designed for stepper drivers (?).

So I see three options open for my design.

1) High quality (Trinamic 2130) closed loop uStepper drive that has 16bit resolution magnetic feedback, but limited to 46V 1.5A rms.

2) High quality (Trinamic 5160) open loop drive that has 4.5A rms 60V. Possibly opening door for Nema 34?

3) Closed loop stoppers and drive such as Leadshine. Unkown quality of drive for me. Are they super smooth and silent? That smoothness of drive reflects in print quality.

10. ## Re: Stepper Optimal RPM Range?

Alright! I went for option 2! I'll have to give up closed loop and trust that well tuned stoppers and drivers will not loose steps.

So now, with 4.5A and 60V what motors would be a good match? Again 20mm ball screws and minimal gantry mass (probably 1kg or less). I suppose I could even go with Nema34 for the gantry motors as they sit on the frame.

11. ## Re: Stepper Optimal RPM Range?

Hi,

So now, with 4.5A and 60V what motors would be a good match?
These have only 0.75mH and may go fast enough for you.

Object reference not set to an instance of an object.

I suppose I could even go with Nema34 for the gantry motors as they sit on the frame.
34 Size have even higher inductance and will go even slower....don't go there.

Craig

12. Originally Posted by joeavaerage
Hi,

These have only 0.75mH and may go fast enough for you.

Object reference not set to an instance of an object.

34 Size have even higher inductance and will go even slower....don't go there.

Craig
Thanks for all your help Craig!

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•