curious about motor speed vs travel speed

[Michael Weber]

Not a cnc user (interesting though) but working with some steppers at work. I can get about 375 rpm (400 microsecond input pulses) before the motors just hum and don't move. I figure using a 12 threads per inch at this rpm gets about a maximum of about 31 inches per minute travel speed. Yet I read about cnc machine jog speeds in the 100's of inches per minute. Are the motors/driver I'm using just very slow compared to those typically used for cnc machines? Or, is there some other drive method that cnc machines use to achieve the high speed movement? The motors/drivers I'm using at work are from Automation Direct. Motors are nema 23's 2.8 amp 276 ounce torque and the driver is set for 400 steps per revolution. The driver can micro-step further but 400 pulses/sec is as slow (fast for the motor) as it gets.

[John Coloccia]

I'm not a CNC guy, but I do have quite a bit of experience working with steppers/drivers and actually designing some drivers also. How are you driving the motor? Are you just telling it to step at 400RPM or are you using an acceleration profile? Nema 23 is a 1.8 degree resolution motor, for 200 full steps. You're taking 400 so you are half stepping. That's not a bad thing. You're only loosing a small amount of torque, but just so you know. Also, just so you know, if you're going to half step, when you power down the stepper will go to a full step....randomly (or not). This is a main source of beginner errors I see. They half step and hold at some odd number of steps.....103, for example. Then they power down. When they start back up, now they're sitting at 102 or 104...103 is actually between steps. The same thing happens with dual phase full stepping. You see this kind of error a lot in low power systems where they depend on detent torque to keep things from moving, and only power the steppers when needed for motion.

There are a million different motor/driver combinations out there to do just about anything you want. You have to pick the right combo for the the application. I don't know that a Nema 23 with your particular driver is or isn't appropriate for a CNC system.

Incidentally, for a given stepper the top speed is mainly determined by the inductance, i.e. how quickly can I ramp up the current in the coils. You have maximum torque holding. As you start stepping, the torque goes down and at a certain point becomes so low that it can no longer overcome friction of the shaft and the motor starts dropping steps. At that point, the motor decelerates and the situation becomes absolutely hopeless....it comes to a screeching halt and sits there buzzing. You won't hurt anything like that, but it's not useful either. That's why I asked about the acceleration profile. You can accelerate hard when you're going slowly, but must accelerate easier as you go faster. Even if the motor/driver combo is capable of much faster speeds, you may never get there if you just start outputting pulses at the target speed or otherwise neglect to use an appropriate acceleration.

[Michael Weber]

Thanks John. I'm using LabView programming software and National Instruments fieldpoint pulse generator to drive the motor. I was simply applying a fixed frequency pulse applied until the motor would no longer turn on from a stopped condition. Right now a 400 microsecond pulse to the driver is the fastest that will start the motor. Monday I'll rewrite the program to start a ramp up in frequency just to see if I can get the motor to turn faster. Not sure why but the driver has no option for full steps but only half steps and up. Seemed kind off strange.

[james mcgrew]

my machines have 5 to 1 gearboxes

[John Coloccia]

Thanks John. I'm using LabView programming software and National Instruments fieldpoint pulse generator to drive the motor. I was simply applying a fixed frequency pulse applied until the motor would no longer turn on from a stopped condition. Right now a 400 microsecond pulse to the driver is the fastest that will start the motor. Monday I'll rewrite the program to start a ramp up in frequency just to see if I can get the motor to turn faster. Not sure why but the driver has no option for full steps but only half steps and up. Seemed kind off strange.

I've done some Labview programming, though not much with steppers. I seem to recall, though, that they have a VI for steppers that will handle things like motion profiles. Also, there is a great paper out there for approximating a continuos motion profile with discreet steps. It's really not that big of a deal...you can figure it out yourself if you really want to, but I ran across the paper some time ago and noted that the implementation details were conveniently geared towards efficient embedded, discreet implementation. If I can dig it up from the caverns of my mind, I will.

Your main problem right now, though, is that you will never drive that motor properly without reasonable motion profiles. It can actually be kind of tricky to calculate the right profile on the fly. I suggest you go down the path of setting up a system where:

1) you have a VI that runs at some fixed, high frequency interval
2) each time through the VI, you decide whether or not it's time to apply a pulse

The trick is to figure out when to start/stop accelerating and decelerating, especially if the movement is short and the acceleration/deceleration portions overlap (i.e. you never get to maximum speed). The paper I'm thinking of had a pretty clean and clever solution, while at the same time optimizing the motion to fit within the frequency of your control loop (it's discreet so you're never exactly on....that needs to be managed).

Best, though, would be to just see if you can find a stepper VI to handle all of the messy details.

[Michael Weber]

Thanks James. Gearboxes, who knew. None of the rudimentary information about cnc that I've read or hobbyist machine plans I've looked at made any mention of gearboxes. Makes sense, but I suppose it takes a more powerful motor than one might find on a basic hobby machine.
John, thanks again. I'll see if I can find a VI on the web tomorrow. Good suggestion on the VI architecture. I'll play around a bit and see what happens

[Angie Orfanedes]

I have Nema 23 steppers on my CNC router with rack and pinion drives, which provide about a 3:1 step-down in speed; see
http://www.cncrouterparts.com/rack-a...a-23-p-50.html

My machine can travel about 1000 ipm, but I find that a bit scarey so I usually keep it below 200 ipm. The motors are driven by this controller, which is a popular one for home hobbyist CNC machines as well as some commercial ones:

http://www.geckodrive.com/geckodrive...ives/g540.html

Look at the bottom of this page for the documentation.

Angie

[Neal Schlee]

I have Nema 23 steppers on my CNC router with rack and pinion drives, which provide about a 3:1 step-down in speed; see
http://www.cncrouterparts.com/rack-a...a-23-p-50.html

My machine can travel about 1000 ipm, but I find that a bit scarey so I usually keep it below 200 ipm. The motors are driven by this controller, which is a popular one for home hobbyist CNC machines as well as some commercial ones:

http://www.geckodrive.com/geckodrive...ives/g540.html

Look at the bottom of this page for the documentation.

Angie

I've rebuilt Multicam R&P transmissions more than a few times. The CNC Router Parts tranny looks to be a very nice and simple solution for R&P stepper systems, reasonably priced to. Just needs a dust cover.
And of course Gecko stuff is good too.

Neal

[Michael Weber]

Thanks Angie. I looked at the specs and noticed a much higher pulse input frequency than I am using. I did manage to speed up the motor rpms above the preiviouis high of 375 by ramping the frequency slowly per John's post. I was just curioius how the cnc machines achieved the high ipm speeds. I read a bit this morning about stepper motor theory but mostly just gave myself a headache. If and when I persue this interest it will certainly be with an available kit.
thanks to all

[George Carlson]

Stepper motors are current driven devices. Because the windings have quite a bit of inductance, it can be difficult to bring the current uip quick enough to get decent torque at high speed (dI/dT). To get around this, drivers use higher voltages with circuits that limit the current. If you see a voltage rating on a stepper motor, it is the voltage required across the winding to get the proper current at stady state. Most small steppers are run using 24V supplies. Supplies up to 48V are common for high proformace systems.
Using microstepping can really help improve torque. With the proper drive/power supply you should be able to run at 3000RPM without any problems.

[John McClanahan]

I have built 2 stepper driven router tables, and I would say that your speed is correct for your setup, and nothing is wrong. Your drivers are set for half stepping, hence the 400 steps per turn. To get more travel speed you need far fewer TPI on the lead screw. Some setups use as few as 2 turns per inch of travel. That would be a 10 TPI, 5 start screw.

Unlike regular motors, steppers produce maximum torque at slow RPMs. As the motors speed up they will loose power, to the point they simply stall and squeal, even with no load. I don't mean to argue, but unlike servos, steppers don't run in the thousands of RPMs. A few hundred is more like it.

Stepper motors have a bad habit of producing a lot of vibration at medium speed. Fractional stepping, or micro stepping helps to reduce this. Micro stepping increases resolution, improving machine accuracy when a low TPI lead screw is used, or with a rack and pinion drive. Micro stepping also requires higher pulse streams from the controller. For older, slower PCs, this can sometimes be a problem, and limit maximum speed.

A good source for information is at geckodrive.com. They have a FAQ section that explains things well. Also check out cnczone.com.

John

[Michael Weber]

Thanks John, good info.