RPM on steppers & StepServos

[A_Camera]

Moving table or gantry makes no difference, mass is mass. Force=Mass * Acceleration It requires the same force to accelerate a 300 Kg table as it does a 300 Kg gantry, or in the case of imitheo's machine, a 30 Kg gantry. From an machine design perspective, it's a little easier to apply the force to a moving table rather than to a moving gantry.

In theory yes. But in reality it makes a huge difference if it is a moving table or moving gantry. And the higher the gantry is the more difficult it becomes. That is one reason why moving gantry design is almost always considerably slower and less rigid and more difficult to make than moving table. Yes, for top speed only the mass and friction is important but for acceleration even the center of gravity.

[imitheo]

Once the gantry has accelerated to speed, the torque requirement drops off rapidly. At that point you are only looking at friction losses, and if cutting, the cutting forces.

If you accelerate too fast, the stepper will decouple internally and growl at you. In other words, the field will be rotating faster than the rotor can keep up with. No damage will occur, but it won't turn again until you reduce the pulse speed so it can re-couple.

I see... that's one my fears, i calculated the torque requirements for my gantry and found a stepper that can keep up to the 1200 RPM real nice with my torque requirements. BUT i fear that when i accelerate too fast it will happen that.

In theory yes. But in reality it makes a huge difference if it is a moving table or moving gantry. And the higher the gantry is the more difficult it becomes. That is one reason why moving gantry design is almost always considerably slower and less rigid and more difficult to make than moving table. Yes, for top speed only the mass and friction is important but for acceleration even the center of gravity.

I'm afraid i'll go for moving gantry as the "compactness" will play a large role. I'll be putting my fourth axis, some custom ATC we are making in the table, so i think it's better that the gantry moves it.

My calculations showed that by reducing some parameters in acceleration e deacceleration the average force that acts on the gantry center of gravity will not be as gore as i thought. Maybe it will work, i'll post my calculations here some day this week so you guys can see if i am mistaken.

[Jim Dawson]

In theory yes. But in reality it makes a huge difference if it is a moving table or moving gantry. And the higher the gantry is the more difficult it becomes. That is one reason why moving gantry design is almost always considerably slower and less rigid and more difficult to make than moving table. Yes, for top speed only the mass and friction is important but for acceleration even the center of gravity.

It depends on the design of the gantry support system, and how the drive system is attached to the gantry. There are several different designs possible, some work better than others.

I don't see where CG is in the F=MA equation.

I see... that's one my fears, i calculated the torque requirements for my gantry and found a stepper that can keep up to the 1200 RPM real nice with my torque requirements. BUT i fear that when i accelerate too fast it will happen that.

I'm afraid i'll go for moving gantry as the "compactness" will play a large role. I'll be putting my fourth axis, some custom ATC we are making in the table, so i think it's better that the gantry moves it.

My calculations showed that by reducing some parameters in acceleration e deacceleration the average force that acts on the gantry center of gravity will not be as gore as i thought. Maybe it will work, i'll post my calculations here some day this week so you guys can see if i am mistaken.

A moving gantry is fine. There are thousands of machines out there with moving gantries.

I don't understand why you need 1200 RPM. A stepper motor is approximately a constant or decreasing horsepower device, they develop about the same output power at 1 RPM as they do at 1000 RPM. This is because the torque drops off as the RPM increases. For an electric motor HP = (Torque(in/lbs) * RPM) / 63025, this is a little different than the automotive HP calculation.

Gear the drive system so that you can operate the motor at lower RPM and thus have more torque available. As I said in a previous post, I am moving a 300Kg load (table) at 15M/min at 570 RPM with a NEMA 34, 1280 oz/in (9.12 N-M). The step resolution is ~0.001 mm (~20000 steps/rev)

[imitheo]

The CG is not in the equation, you relate the average force with the change of momentum that the machine is suffering when deaccelerating to the arm of the force on the CG of the gantry. I THINK this is the way to see if the structure it's going to stay in place.

I need 1200 RPM because my ballscrew has 10mm lead, and i need 12m/min.

[ger21]

Use a 2:1 belt drive to run the motor at 600rpm.

[Jim Dawson]

The CG is not in the equation, you relate the average force with the change of momentum that the machine is suffering when deaccelerating to the arm of the force on the CG of the gantry. I THINK this is the way to see if the structure it's going to stay in place.

It really depends on how the gantry is supported on the rails. The twisting moment may or may not have effect on the overall system.

I need 1200 RPM because my ballscrew has 10mm lead, and i need 12m/min.

A 1:2 ratio would put the motor at 600 RPM at 12m/min. With a 35 Kg gantry even a 1:3 ratio would work, thus further reducing the motor RPM. I have found that running stepper motor slower is better, and you can still maintain machine resolution. Just for fun, look at your calculations again using lower motor speeds.

EDIT: Gerry you beat me to it

[NIC 77]

What kind of competition is this? High school, college, university? What kind of machine is this? A router? What kind of spindle will you have on it? What are the other parameters you will be tested on?

The only thing they state it, is that it needs to pass through the sensors with 12m/min and the machine will have a 75mm to accelerate it to that speed.

so 12m / min = 472 IPM. This is just a rapid speed? No cutting required at this speed? Acceleration is what limits speed (keeping ballscrew lead and motor performance constant). There are a few types of control software that have a separate acceleration input for rapids. Flashcut, for example. Mach 3 does not, meaning that the same acceleration you use for your rapids is what you have for your feeds. An acceleration of 0.035G will get you to 500 IPM in 65mm. This value is very low, meaning I can't see your machine actually cutting shapes fast with it. That's where the test should focus.

Your Z axis also needs to hit 12m / min?

You really need to find torque vs speed graph for the motors! Ask the motor manufacturer if it's not online. This is where you need to start. Also, you need to use the right amp driver and right voltage power supply. Post that info, along with the outer diameter of the ballscrews you're using, the lead you already said was 10mm, the lengths of your screws, the weights of your moving parts, ie: entire gantry, (includes everything including z carriage and z up and down part), Z carriage (includes z up and down part), Z up and down part. Basically all the weights that each screw will move. Also see if you can find the rotor inertia for the motors. I can throw it in my spreadsheet and see what it says. We can compare math.

A 2:1 belt reduction system should work, but with the low accelerations required, you probably don't need it. Still, you need to do the math. Also, you need to be careful about using higher reductions and planetary gears if you want good acceleration performance. The motor needs to accelerate faster than the ballscrew by a factor equal to the gear ratio when you use reduction. It doesn't sound like fast acceleration is really your goal though.

[NIC 77]

I found this article online:

Improving High-Speed Performance of Hybrid Stepper Motors

Of course you'd need to find a baseline graph for the motors you want to use.

On a bit of a side note, I realize that you can buy an 8 wire motor and wire it in series or parallel, but has anyone ever taken a 4 wire motor and opened it up and been able to change it to a parallel configuration? Assuming of course, that it was wired in series to begin with.

I'm guessing that those stepper motors that have feedback encoders and go to the faster RPMs are already wired in parallel even if they only have four wires coming out of the motor?

[A_Camera]

The CG is not in the equation, you relate the average force with the change of momentum that the machine is suffering when deaccelerating to the arm of the force on the CG of the gantry. I THINK this is the way to see if the structure it's going to stay in place.

No, the CG is not part of the speed, but acceleration/deceleration possibilities are worse with high CG and since you can't lose steps or use infinite acceleration and also not running in vacuum, CG is a very important factor. CG doesn't matter for a moving table but the gigher the moving gantry is the more problems you will experience. Just like a low sports car is more efficient in acceleration than another with the same engine, tyres and mass if that other one is considerable higher. Anyone who ever transported a large object on their car roof knows this. The top speed may be the same in the end but it will take much longer to accelerate (not only because of drag) and decelerate, and also the whole journey becomes more unstable, turns insecure (must be taken slower) and so on than if the same car had the same mass placed in the boot and the passenger compartment with lower CG. That's why I am saying "in theory".

If you have a limited acceleration/deceleration distance, even though in your case that 70mm is HUGE, you must of course try to keep the CG low. One way of doing it is that if you add some extra weight at the bottom, where the ball nut is. It will increase the mass but will move the CG down. Just an idea you might want to try when you encounter speed problems.

I need 1200 RPM because my ballscrew has 10mm lead, and i need 12m/min.

Like I said, with my 1605 and NEMA23 I can get that speed with my moving table machine. So if the CG does not matter, only the mass, then with the 10mm pitch you have you should get twice as much, which of course at least I don't believe in at all, but if you are so sure about that CG doesn't matter then you should not worry at all. 12m/min should be an easy match...

One thing I don't remember was mentioned is the motion controller and the software you are planning on using. I think those two play an important role as well. I don't think you can get that high speed with parallel port use, but I can be wrong about this as well...

Using a belt reduces the needed rpm and that in turn reduces the number of steps needed for the speed, and may reduce the need of quality electronics and software as well.

[NIC 77]

[QUOTE=A_Camera;2060422
Like I said, with my 1605 and NEMA23 I can get that speed with my moving table machine. [/QUOTE]

That's 2400 RPM. Fast for a stepper motor. Please forgive my curiosity, but what kind of Nema 23 are you using, how heavy is your table and what are your acceleration settings if I may ask?

[A_Camera]

That's 2400 RPM. Fast for a stepper motor. Please forgive my curiosity, but what kind of Nema 23 are you using, how heavy is your table and what are your acceleration settings if I may ask?

https://adapting-camera.blogspot.se/...nt-status.html

Sorry, I was wrong. 10m/min is my maximum, but even that is more than 1200rpm and if I can get that speed with my 1605s then he should easily be able to get 12m/min with the 10mm pitch screws he is using, especially since some of you claim that moving gantry or moving table makes no difference. My table is 17.5kg without the two vices and the large 10mm aluminium plate plus the PCB fixture I have on permanently, so total table weight, when in use, is about 30kg. His gantry is 35kg, so again, very similar to my table in terms of weight. Specifications for my machine are in the above link. Z weight without stepper and spindle. On my blog there are many other details about my machine and the different stages of building/upgrading. I also have many YT videos.

[NIC 77]

Sorry, I was wrong. 10m/min is my maximum, but even that is more than 1200rpm and if I can get that speed with my 1605s then he should easily be able to get 12m/min with the 10mm pitch screws he is using.

So you could get 10m/min with about 0.1G acceleration using only one motor and 5mm lead ballscrews. That's impressive.

Are you using the jss-motor 57HM56-3004??

Changzhou Jinsanshi Mechatronics Co. Ltd.

I couldn't find a torque vs speed graph for it. 0.9 deg/step instead of 1.8. How is that wired inside? Parallel?

If we could find a torque vs speed chart for it, the OP could determine it's use for 10mm lead. Perhaps that would solve his problem right there. If that's what you're using?

[imitheo]

I found this article online:

Improving High-Speed Performance of Hybrid Stepper Motors

Of course you'd need to find a baseline graph for the motors you want to use.

On a bit of a side note, I realize that you can buy an 8 wire motor and wire it in series or parallel, but has anyone ever taken a 4 wire motor and opened it up and been able to change it to a parallel configuration? Assuming of course, that it was wired in series to begin with.

I'm guessing that those stepper motors that have feedback encoders and go to the faster RPMs are already wired in parallel even if they only have four wires coming out of the motor?

That's a really nice article, thanks a lot for it. I knew that increasing voltage would make the motor run faster, but it's good to see graphics and stuff.

If you have a limited acceleration/deceleration distance, even though in your case that 70mm is HUGE, you must of course try to keep the CG low. One way of doing it is that if you add some extra weight at the bottom, where the ball nut is. It will increase the mass but will move the CG down. Just an idea you might want to try when you encounter speed problems.

Yes i was calculating the momentum the structure would suffer when deaccelerating or accelerating and it came to me that this 70mm is VERY HUGE. So i am not worrying about my machine base right now. Maybe the vibrations, but i'll be a Ø6 endmill so no heavy milling will be done.

So you could get 10m/min with about 0.1G acceleration using only one motor and 5mm lead ballscrews. That's impressive.


Are you using the jss-motor 57HM56-3004??


Changzhou Jinsanshi Mechatronics Co. Ltd.


I couldn't find a torque vs speed graph for it. 0.9 deg/step instead of 1.8. How is that wired inside? Parallel?


If we could find a torque vs speed chart for it, the OP could determine it's use for 10mm lead. Perhaps that would solve his problem right there. If that's what you're using?

Yes, pretty impressive, i'm starting to feel really confident about my motors and my design. In fact i would like to point that there is a big difference in
table and gantry design, just like A_camera stated, the CG of the gantry will be really high, so any stops and starts will cause a moment with the arm of the gantry CG.
The table has a really low CG, so the moment is almost nothing compared to the gantry moment. I am well aware of that, but i'll still go for my gantry design because it will be small.

And another factor that goes for the gantry is: I'll be using my machine surrounding to make an enclosure. In this competition, we have 12 hours to make the machine, BUT as 12 hours is little time, we can take the structure mounted, but other stuff, like electronics, enclosures and a custom ATC UNIT is demanded to be made there, machined i mean. I'll be personally handling the lathe and the milling center(DISCOVERY 760).

[imitheo]

Have anyone ever look at the TBI MOTION ballscrew designing PDF ?

TBI MOTION Ballscrew

I used this design guide to know the Torque i would need to drive my gantry. If anyone could take a look with for me to be more confident about these calculations.

To get the cutting force on my endmill i used this KENNAMETAL calculator:

KENNAMETAL Calc

To see the maximum cutting force my end mill could hold, i got the milling parameters for the endmill i am using from kennametal and then inserted in the calculator.

So with the maximum tangential force i maxed my parameters using my spindle power. Finally i got 90N as my cutting force.

Anyone could take a look at these ?

[A_Camera]

So you could get 10m/min with about 0.1G acceleration using only one motor and 5mm lead ballscrews. That's impressive.

Are you using the jss-motor 57HM56-3004??

Changzhou Jinsanshi Mechatronics Co. Ltd.

I couldn't find a torque vs speed graph for it. 0.9 deg/step instead of 1.8. How is that wired inside? Parallel?

If we could find a torque vs speed chart for it, the OP could determine it's use for 10mm lead. Perhaps that would solve his problem right there. If that's what you're using?

My motors are bought four years ago and I don't think you can buy those today. They are called 57BYGH633 and sold by Wantai Motors. I have no idea about any other data than what's stated by the documentation.

Here is a link I found.

[NIC 77]

My motors are bought four years ago and I don't think you can buy those today. They are called 57BYGH633 and sold by Wantai Motors. I have no idea about any other data than what's stated by the documentation.

Here is a link I found.

Thanks for the info. Some of those torque graphs are almost flat! Not usually what you see with a stepper motor. I'd like to see if there is a Nema 34 motor that is almost flat up to 1300 RPM at about 3.5 Nm. Yes, I could use a servo, but it would be easier to find a flatish type stepper for my own purposes at this point seeing as I already have Gecko drives, power cables, breakout board, etc, for stepper motors.

[NIC 77]

Have anyone ever look at the TBI MOTION ballscrew designing PDF ?

TBI MOTION Ballscrew

I used this design guide to know the Torque i would need to drive my gantry. If anyone could take a look with for me to be more confident about these calculations.

To get the cutting force on my endmill i used this KENNAMETAL calculator:

KENNAMETAL Calc

To see the maximum cutting force my end mill could hold, i got the milling parameters for the endmill i am using from kennametal and then inserted in the calculator.

So with the maximum tangential force i maxed my parameters using my spindle power. Finally i got 90N as my cutting force.

Anyone could take a look at these ?

Yes, take a look at my thread here:

http://www.cnczone.com/forums/linear...118-cnc-3.html

Look at post #21. There is a sample calculation.

You can try to recreate my graph in post #5 using an excel spreadsheet or whatever kind of math program you want to use. Then if you get the same results, both of us will know that our spreadsheets are good and you can then put your specific parameters in your spreadsheet for your machine

The motor torque table is here:
http://www.cncrouterparts.com/cdata/...rque_curve.pdf

That is two motors driving an 800lb gantry at 0.25G, And calculating how much force is left over at different RPMs. I didn't include preload drag or bearing drag and assumed an efficiency of 90%. See if you can recreate that graph.

Some references are here:

http://www.engr.sjsu.edu/bjfurman/co...iaformulas.pdf
https://www.aerotech.com/media/85449...d%20sizing.pdf
http://www.hiwin.com/pdf/ballscrews.pdf pages 26 and 27

Also, HSMAdvisor gives cutting forces and will tell you how much HP you need. Your machine is light and I doubt (correct me if I'm wrong) that you will have a really powerful spindle, so I don't think you will be making really aggressive cuts or subsequently need much cutting force.

But you can check it out and download it here if you wish:

http://hsmadvisor.com/?page=HSMAdvisor

Post your results when you're done, let's see what you come up with.

[imitheo]

Hi again, long time no posts, i was testing some stuff with my steppers. Anyway i didn't had the time to make the calculations because my team is all over me, okay, i have some questions regarding the speed that i noticed under my tests and i would like to share here:

First of all, the parameters and the motor:
KTC-HT23-401 Datasheet
Running on 48v PSU
I am using a frequency generator to generate the square wave to pulse the motor.

1) Okay, i first tried to run at full step and i noticed that it was really hard to accelerate it really fast, the motor was suddenly stopping all the time. Then i raised the resolution to 1/4ustep and bam, it was so easy to accelerate and deaccelerate it, anyone can explain me why ? And the vibration also lowered.

2) I was trying to measure the current in one phase to get the total average current, so i set my multimeter to AC and it's a True RMS one, and measured it. The values that show on the display(RMS) and the values that datasheet shows are the same ? Or the value in the datasheet is the peak current ?

3) Also that value that shows on the driver, what current is this? Average, Peak?



4) Another question regarding the driver, the motor i am using shows 4.3A/phase, so i bought a driver that runs 4.5A, ON MY UNDERSTANDING, as they sell the motors with CURRENT/PHASE, the AMPERAGE rating on the DRIVER, ALSO WAS PER PHASE. BUT i am doubting this.

The driver i bought --> 4.5A Action Motors Drivers

Why i doubt this ? I set my driver to 4.5A, and i expected to see on my multimeter display, 4.3~4.6A RMS, but i'm getting 2.2A on my measured phase. SO am i correct? The amperage rating on the drivers is for the total current that the driver can handle, and the vendors rate their motors with A/Phase ?

And i should get this bulky, 8A driver ?
Bulky Hardened Super Hulk 8A Driver

5) (ASSUMING EVERYTHING I ASKED BEFORE IS CORRECT)My motor is getting pretty hot with 48V 2.2A/Phase, and it's rated for 4.3A, if i get my bulky driver and set for 8A max, this thing is going to fry ! Some fans around it should do the job ?

[NIC 77]

4) Another question regarding the driver, the motor i am using shows 4.3A/phase, so i bought a driver that runs 4.5A, ON MY UNDERSTANDING, as they sell the motors with CURRENT/PHASE, the AMPERAGE rating on the DRIVER, ALSO WAS PER PHASE. BUT i am doubting this.

https://www.linengineering.com/resou...-vs-amps-peak/

"Amps/Phase * 1.41 = Amps Peak current. Regardless of whether you remember the reason behind the 1.41 value, it’s crucial to understand this relationship because in most manufacturers, drivers only care about what peak value it can output. And step motors only care about listing the Amps/Phase value. As long as you understand what the difference is, you will be able to talk in the same language to both driver manufacturers and step motor manufacturers."

That is a quote from the article I posted. That's according to Lin Engineering. You can tell me if this is correct, I don't know.

And someone can correct me if they disagree with these two statements:

Also, the motor will not draw full amps unless running at full torque and low speed.

The stepper motor can only draw fewer amps at higher speeds before it starts to loose steps, and has less power, and less torque available. This is different from a spindle motor that has more power at higher speeds, but also less torque.

5) (ASSUMING EVERYTHING I ASKED BEFORE IS CORRECT)My motor is getting pretty hot with 48V 2.2A/Phase, and it's rated for 4.3A, if i get my bulky driver and set for 8A max, this thing is going to fry ! Some fans around it should do the job ?

You'd have to limit the current with the switches on the driver. No, you can't run that motor with 8 Amps. That's 4.24 amps, and if you believe Lin Engineering, that's 6 Amps max peak current, but before doing that I'd recommend having a very good look at the driver manual to determine what they are talking about and what their current settings actually mean. It won't help you get better performance at faster speeds as a stepper motor uses less amps at faster speeds. Please do some tests to verify if what I've said is true. I'd like to hear about the results. You'd have to test it under load.

Increased voltage gives better performance at faster speeds.

For the heating, what is the insulation class of the wires on that stepper? Also, I'm guessing that a Nema 23 can't dissipate heat as well as a Nema 34? If you heat sink your motors with a fan, well, I've not heart of it on a CNC router, but it would be interesting to see the results.

NEMA Insulation Classes

[imitheo]

https://www.linengineering.com/resou...-vs-amps-peak/

"Amps/Phase * 1.41 = Amps Peak current. Regardless of whether you remember the reason behind the 1.41 value, it’s crucial to understand this relationship because in most manufacturers, drivers only care about what peak value it can output. And step motors only care about listing the Amps/Phase value. As long as you understand what the difference is, you will be able to talk in the same language to both driver manufacturers and step motor manufacturers."

That is a quote from the article I posted. That's according to Lin Engineering. You can tell me if this is correct, I don't know.

And someone can correct me if they disagree with these two statements:

Also, the motor will not draw full amps unless running at full torque and low speed.

The stepper motor can only draw fewer amps at higher speeds before it starts to loose steps, and has less power, and less torque available. This is different from a spindle motor that has more power at higher speeds, but also less torque.



You'd have to limit the current with the switches on the driver. No, you can't run that motor with 8 Amps. That's 4.24 amps, and if you believe Lin Engineering, that's 6 Amps max peak current, but before doing that I'd recommend having a very good look at the driver manual to determine what they are talking about and what their current settings actually mean. It won't help you get better performance at faster speeds as a stepper motor uses less amps at faster speeds. Please do some tests to verify if what I've said is true. I'd like to hear about the results. You'd have to test it under load.

Increased voltage gives better performance at faster speeds.

For the heating, what is the insulation class of the wires on that stepper? Also, I'm guessing that a Nema 23 can't dissipate heat as well as a Nema 34? If you heat sink your motors with a fan, well, I've not heart of it on a CNC router, but it would be interesting to see the results.

NEMA Insulation Classes

As i expected, the drivers rate the total peak current, so the stepper is rated on amps/phase, actually amps rms/phase. So my motor, that is rated 4.3A/Phase, should be run at a driver that is capable of reaching 12A. THATS A LOT :/

My team is probably giving up on this stepper, it may reach our purpose but a 12A driver is insane and the price reaches the price of a 200W Servo Motor. It may not need this amount of Amperage for our machine, but we don't have the money or the time to start testing stuff, buy a expensive driver, buy a PSU, everything. We probably go for a more safe option.

IF we try the tests and i heatsink it with a FAN, i'll inform you guys.