Fastest stepper driver?

I'm looking for a very fast stepper driver, beyond any printable speed, something pushing 500mm/s or more.

I'm designing a printer with 10:1 geared X Y steppers so I would like to be able to push it to a 50mm/s print speed.

Edited 1 time(s). Last edit at 03/21/2013 12:35PM by Dirty Steve.

... I have some IMS stepper drivers capable of 1/256 microstepping (51200 steps per turn with a 200steps-motor) with up to 10MHz step rate - something like 11718 rpm max --- but haven't verified this speeds ...

Or Precitec/Berger&Lahr drivers and motors with 1/10 microstepping with 10000 steps per turn and 200kHz stepping rate - 'only' 1200 rpm's, but tested/verified with 2Nm steppers and 5mm-pitch spindles to 100mm/s traveling speed with 0.5 microns resolution

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Our stepper drivers (out of stock at the moment, hopefully they will be in shortly) from Panucatt have no problem to over 1M/s so you would have plenty of headroom to gear them down. Roy @ [www.panucatt.com] may have some in stock.

[store.qu-bd.com]

Chelsea - QU-BD

@Chelsea

Those are what I'm looking for, looking into running a Azteeg X3 with a Raspberry Pi.

Chelsea - QU-BD Wrote:
-------------------------------------------------------
> Our stepper drivers (out of stock at the moment,
> hopefully they will be in shortly) from Panucatt
> have no problem to over 1M/s so you would have
> plenty of headroom to gear them down. Roy @
> [www.panucatt.com] may have some in stock.
>
> [store.qu-bd.com]

According to the datasheet for the DRV8825, its max step frequency is 250kHz, which is actually slower than the A4988. Where did you get the 1MHz figure from?

There is no way an AVR will drive the stepper near 1MHz while also doing speed ramp etc anyway.

UH, I'm completely clueless, but,
Isn't movement speed increased by more amperage so that you can have higher toque thus large diameter pulley?
Lower impedance/higher voltage to switch faster?

And mostly, wouldn't 1/Xths stepping (However ridiculous and unlikely that is to be meaningful under a load) require more steps, slowing you down?

They are fine to move at 1Meter per second which is double the capability you were looking for. We have tested the 4988 and they simply don't work for our setup at high speed.

Chelsea - QU-BD

You can run Nema17 motors at ~1000RPM with simple pololu drivers. That's IME about as fast as you can run steppers reliably even with expensive drivers.
You can't easilly do that with uStepping enabled on arduino based electronics because it requires the interrupt rate be too high, I did some tests over the weekend and it's relatively trivial with a 100MHz Arm board so something like smoothie ought to b able to do it.
Repetier firmware on an arduino might get you close if you set quadstepping, but the jitter might be enough to stall the motor at the limits.
Worth noting that at those speeds steppers have greatly reduced torque and you might need a higher input voltage and more current to drive the printer which might eliminate the pololus.

IMO your best bet to get higher speeds than 500mm/s is to use bigger pulleys and trade off speed for accuracy and torque.

Polygonhell Wrote:
-------------------------------------------------------
> IMO your best bet to get higher speeds than
> 500mm/s is to use bigger pulleys and trade off
> speed for accuracy and torque.


I agree, also, you are going to use exponentially more torque the faster you step, so going to twice the torque and double pulley size could mean several times more practical torque than you might expect.

Also don't forget to ramp into (Accel/jerk) carefully. You are right, all these things call for an ARM : /. I only worked with ATmega until now but it seems like the DUE will get more and more traction. For me, right now it is 5X the cost.

any one think about gecko controllers?

I have a few gecko 201 stepper drivers and have spun decent square body nema23 steppers over 2500rpm. There is hardly any torque at that speed. The gecko are 10 microstep drivers. That's over 83,000 steps per second. Power supply was set at 48volts. You will be able to spin the stepper faster by using a 80volt power supply which is the max rated voltage of the geckodrive. My lab supply is only 55volts maximum. I used a Tektronix frequency generator to output a smooth jitter free pulse train. Can a arduino output a jitter free pulse that fast while doing simultaneous 3 or 4 axis moves?????

Using good stepper drivers and motors, you should be able to run a stepper reliably at 1000rpm with decent available torque. The size of your gearing, belt drive/pulleys etc. will determine your maximum axis speed. The limitation will most likely be the arduino.

Jim Fong Wrote:
-------------------------------------------------------
>
> Using good stepper drivers and motors, you should
> be able to run a stepper reliably at 1000rpm with
> decent available torque. The size of your gearing,
> belt drive/pulleys etc. will determine your
> maximum axis speed. The limitation will most
> likely be the arduino.

I guess "decent available torque" depends on what torque you start with, but the torque curve on stepper motors drops very rapidly. A typical example:



The torque at 17 RPS (1000 RPM) is about 10% of the max torque. You might find some motors with a slightly better performance, but the fact is, stepper motors are just not designed to be run at speed.

With a perfect driver, the limitation is really the motor, but a poor driver implementation may get even less performance out of the motor.

Bobc, I agree that steppers do loose most of their torque at higher speeds but choosing the right drivers, steppers, linear components makes all the difference. I use a cnc router that rapids at 200ipm with 5tpi ballscrews. The stepper spins at 1000rpm or more. To say that steppers are not designed to run at that speed isn't always true. The steppers motors are nema23 around 300in/oz.

I believe some of the faster 3d printers are spinning nema17 steppers at really high rpm. Chelsea posted hers can do 1meter/sec!!! Would Like to know what size pulleys and belts she is using to get that kind of speed.

My cnc mill with 800 in/oz steppers 5mm pitch ball screws and 69 volt power supply to leadshine drivers moves @ 236 in/min @ 3200 steps per revolution.

Well steppers may lose most of the torque at high speed, but you mostly need torque when accelerating so that is not a so big deal at constant speed.

However if you want to run that fast, accelerations ramps are mandatory are you will need something like grbl on a more capable hardware than arduino /me thinks.

Stepper motor test

Just for kicks I hooked up a spare stepper motor to a gecko202 driver just to see how fast I can spin the motor. I only used a 20volt power supply since I didn't want to pull the bigger power supply hooked up to my cnc in the basement. The maximum pulse rate from the tek frequency generator is about 91Khz before the stepper stops spinning. That calculates to be about 2700rpm. All this really shows is how good a gecko stepper driver works with only a 20 volt motor supply. Of course everything changes when you have a mass to move with acceleration profiles.

At 1000rpm, you cannot stop the stepper motor with your fingers. On my cnc router you have to really lean hard on the axis to stop it from moving. I usually cut around 50ipm and you can easily break small end mills if it hits a metal hold down.

To get back on topic of 3d printer, I'm using the cnc to mill out some parts for my 3d printer that I am currently building. It's based on the mendel90 design. I have a bunch of spare stepper motors, parker compumoter e-dc stepper drivers, thk linear rails and arduino board sitting unused. I did have to purchase some gt2 20 tooth pulleys and belts though. I haven't decided to build or buy my hot end. I also have a cnc metal bench mill and manual mini lathe that I use to mill out parts. I hope to have it done soon but I work 7 days so finding time in the basement workshop is tough.

Keling nema23 3amp 2.0Nm stepper motor
Gecko 202 10 microstep driver
20volt power supply
Stepper pulse rate 91Khz
Stepper motor spinning at 2700rpm