High Speed Stepper Motor

[RocketNut]

I am in need of a stepper motor with a high RPM ( 750 or higher ). The motors I have been testing tops out at about 300 RPMS.

So the question is there a motor out there that can achieve this speed?

I can not use a normal motor because the shaft only rotates about 90 Deg to 300 Deg then back to the start position at different speeds..

I am using DRV8825 driver chip, controlled with a Raspberry 3.3B and Python.

 

[GIOV]

Hi May the force be with you
Try Find the correct motor in tthe following webs:

Brushless DC Motors | BLDC Motors from Nanotec

Brushless DC motors from Nanotec. ✓Higher efficiency and power density ✓Quiet ✓BLDC motors with encoder, gearbox, controller /drive options. Quote >>

us.nanotec.com

Orbray Co., Ltd.

We are a manufacturer of precision jewel parts, fiber optic components, DC coreless motors and motor units. We have received high acclaim in a diverse range of niche markets, including telecommunications, semiconductors, robotics, precision equipment and biomedical science.

orbray.com

Servo Motors/Drivers/Controllers | TAMAGAWA SEIKI CO.,LTD.

We support various motion controls such as industrial robots, peripheral devices, and assembly equipment, which are essential for factory automation and labor saving with a variety of run-ups and excellent functions. We respond to wide range of needs with different variations.

www.tamagawa-seiki.com

https://www.ti.com/lit/ds/symlink/drv8825.pdf?ts=1691627291423&ref_url=https%253A%252F%252Fwww.google.com%252F

 

[tyc]

If you have surplus torque with your current stepper motor, you could try using a gearbox or pulley/belt mechanism to double your output speed.
If the motor torque isn't enough, you could go for bigger frame motor, and use the the gearbox or pulley/belt trick above.

As GIOV suggested, check brushless DC motors and field oriented control(FOC) method, you would need specific drivers though. Check Makerbase simpleFOC drivers for that. There are many on the market, prices ranging from affordable to less affordable.

check following video for reference.

 

[RocketNut]

Thank you for the great video.

I only need to rotate to 90 Deg or 180 Deg at different speeds. The stepper motor drive can handle up to 250000 ( 250 KHz ) pulses per second. Which should be fast enough. Then return back to a home position (an ITR9608-F Photointerrupt ). The motor controller in the video looks like a lot more than I needed

Maybe I should mention. This is part of a Rasberry running a Python app with an Arduino stepper motor logic controller.

 

[tyc]

Yes, I understand your concerns regarding the complexity of FOC.

How about the load you need to move? I mentioned using speed increasing gearbox or pulley/belt. Are these doable for your application? You would basically increase your output speed, sacrificing your torque output.

 

[RocketNut]

The load mass is less than 1.5 LBS, Gears, and pulley/belt will not be possible, because of the space restrictions. Yes, you are correct the step speed will be different for the 10 settings.

As for the FOC, I would need a lot of help to write the Arduino ( ItsyBitsy 32u4 ) code. I never have written code that has analog outputs. What is the code to rotate to the 90 DEG at different speeds?

Lastly what single cannel drive would you use?

 

[manton]

Your max motor speed is limited by the supply voltage, so the higher the voltage you can run, the faster the motor can spin. It comes down to the back EMF voltage that the motor produces when running fast. If the BEMF voltage reaches the supply voltage, then you have no torque, so the only option then is to increase the voltage. You should be able to find some speed vs voltage graphs somewhere.

 

[JTB]

Hi RocketNut,
Your need are very special:
assuming that the stepper is 200 steps/rev. Which very common. At 750 rpm you will need 2500 steps/s which is very high for std stepper.
I don’t know the required torque at that speed but for sure you will have loses due to the high frequency and also you will need to get the pull-in pull-out torque curves to check that your working point is bellow the curves!
I can understand that the stepper principle is ”simple” to implement in terms of electronic and codes, now for your application I’m not sure it is the best choice. Probably a BLDC or BLAC more would be convinient for dynamic and efficiency. If you can share more info, I will be happy to support you in terms of recommandation or path to follow.

 

[VoltsAndBolts]

This sounds like a job for a Servo motor to me.

 

[JTB]

Hi VoltsAndBolts,
Probably and probably not
By experience, if you don’t excatly the problematic ( needs and contraints), it is difficult to give the right solution. From my understanding RocketNut’s application need direct-dive solution for the dynamic. BUT there is lack of information to say it yet

 

[RocketNut]

Thanks for the great info and for sharing your wisdom with me.

I am not looking for high-speed RPMs pre-say. I only need to swing from the home position (20 deg ) to the 90-degree or 180-degree position and then return back to the home position. This works out to be, with 4 micro-stepping resulting in 0.45 degs per step, 156 (90deg). and 708 (180deg) steps. With the stepping frequency ranging from 750 Hrz ( 0.0013333-seconds pulse width) to 3000 Hrz.(0.0003333-second pulse width ) The motor driver can handle up to 250KHrz.

​

 

[JTB]

Hi RocketNut,
First of all:
Reffering to your first post, the speed needed is 750 RPM ! Equivalent to 2500 Steps/s
Now from your last post, you are request is more about 750 steps/s up to 3000 steps/s, is there a confusion between RPM’s and steps/s?
Just for my understanding of your application, you have two strokes 70° (90-20) and 160° (180-20)
What is the travel time requested for each stroke? Just to confirm the speed and steps/s.
Second:
I agree that the driver can achieve 250 kHz stepping, it is the limit of the driver it doesn’t mean that the stepper can follow such frequency! For sure not at all. It is really important to get the torque curves of the selected stepper (pull-in and pull-out) and to confirm that the required torque is at least bellow the pull-out torque!

Third:
Using micro-stepping helps to get smooth motion, now the drawback of that is that you lose some torque margin, and the higher the inductance of the stepper is the more torque you lose!

As I like to same, motor sizing is also a trade-off of many constraints and the electronic is part of it.
That’s my two cents

 

[RocketNut]

Yes, was a very bad choice of wording in my first posts No I fully understand the difference between RPMs and steps per second.

Yes, that's correct two strokes 70* and 160*.

I really do not know the travel times which will be determined through experimentation. The 750 steps/second is just a starting point.

Here is the datasheet for the stepper we are thinking about using for this project. If I understand the datasheet properly the motor should be able to handle 3000 steps/second. We will be using 24VDC to power the motor.

 

[JTB]

Ok now We start to move!
Regarding the data sheet, we only see the pull out (maximum torque without acceleration), the pull in is not plotted and most of the time you will see a huge difference between the curves and additionnally if your mechanism has a high inertia relative to the rotor this will get worst. So for your application you should generate an acceleration ramp to consider the inertia effect.
Now i don’t know that stepper brand but I have some doubts regarding the torque shape and values.
Just small math:
the electrical time constant is about 2 ms (time to reach 63% of the rated current or voltage limitaion/R) so at 1000 steps/s the step duration is half of that value which means that current reached will be only half or a third of the rated value so the torque should be only half or third of the max pull out.
Second doubt:
The torque constant is assumed to be 0.218 Nm/A which is equivalent to the Bemf constant 0.218 V/(rad/s) equal to 1.36 V/(rev/s) or 1.36 V/(200 steps/s) so at 1000 steps/s, one will get 11.5 V of Bemf and at 5000 steps/s it should be 57.5 V much more that the 24V of the driver in their data sheet.
Either my math are wrong (it is late) either they over estimate their motor performance.
From the electrical time constant, I beleive that the stepper can be used to a speed range up to 500 steps/s or max. 1000 steps/s

 

[JTB]

Additionnal info,
If i read well the data sheet, they talk about half step, which means they have 400 steps/rev which means that the conversion PPS / rev should be ratio 400 instead of 200. But even that doesn’t explain their curve shape.

 

[RocketNut]

The stepper motor we are looking at is the "Pololu Stepper Motor"

Do you have a motor in mind that will fit our needs?

 

[JTB]

Not really,
First of all you need to define more precisely:
The required torque of your application (frictions, gear efficiency if any, torque required for the load displacement and acceleration torque)
Then you need to define the motion profile (accelaration ramp, max speed reached, time to reach the strokes)
You will get then the max speed, the working points and also the required mechanical power
if your intention is to use the TI driver, then you will see if the mech. Power match to the elec. Power at the driver output (take some margin). Otherwise you should search a compatible driver.
For example the stepper motor you selected can have higher speed if your driver is 48 Vdc, the issue will be the efficiency of the system

From that you can start to investigate stepper suppliers