I have an Open Builds Nema 23 High Torque that stalls/skips steps once it reaches ~450 rpm. Is this normal? Perhaps I'm reaching a resonance frequency? Perhaps this is too fast for the stepper? I'm running a 24V power supply and Buildbotics controller. I've run it moving a load via a belt and also free to spin and it always stalls around 450 rpm. Acceleration and jerk settings don't influence the RPM at which the motor stalls. I'm using the motor to move an XL Gantry cart back and forth on CBeam for a lightweight, high speed application (not machining) I get up to 26m/min. Any insight would be appreciated as I'd like to figure out if a different motor may help me speed things up. Thanks!
Try using a 36V power supply. The higher your voltage, the more usable magnetic field (ie. torque) you can build at each step before the rotor moves on. If you have microsteps set high, try knocking them down as well. At those speeds, you a) don't need them, b) are losing torque because of them (because of the above) and c) might be maxing out the pulse generation capacity of your microcontroller?
Rob is far more experienced with these things and will have a far better idea of what might be wrong than I would with your steppers as torque does drop off the higher your speed but... Your stall problem won't be your Buildbotics controller as it has a max output of 250,000 steps per second and for example even an Arduino Uno with Grbl on it can pulse at 40,000 steps per second. Assuming you are using full steps (200) with no microstepping for your 450 rpm that equates to 1500 steps per second so it is below the max steps per second that your controller can output. How long is your gantry and what is your acceleration/deceleration rate and max speed settings? Have you got enough room on your gantry to accelerate, reach your target max speed and decelerate? Out of interest how are you measuring your rpm? Cheers Adam
Yeah, he'd have to be at 256x(!!!) microstepping to max out the chip then. Good info! Still better to be at full steps if possible, or 2x or 4x if it's too noisy. They're a torque killer, for sure, and it falls off a cliff after 8x. Interesting though that acceleration settings don't change anything. Usually it's in the ramp-up that the motor stalls out.
Hello Rob, Do you know where I can find the torque curve for the OB high torque Nema 23? I've Googled it and tried the site search but I can't seem to find it anywhere. In your reply about "acceleration settings don't change anything" is that for Grbl or the Buildbotics firmware? I've been wondering how Grbl works out its acceleration profiles or are they just constant? Cheers Adam
I'd just look up the torque curve of a similarly spec'd (torque, length) NEMA23 motor, tbh. The mechanics are pretty universal at this point and it should give you a decent idea. Grbl's only acceleration control is rate, there's no jerk setting in there that I've seen. At least, after installation. Prior to compilation, there may be some way of adjusting it, but I've never had an issue where I needed to consider it. Other controllers do have jerk settings, which may be in the BB firmware since the OP mentioned it, but I'm unfamiliar.
I went off and did some digging and found a torque curve calculator that creates a predicted torque graph which Ryan Carlyle had kindly posted on Github: rcarlyle/StepperSim You can plug in your numbers for your motor and drivers (I'm not sure what they are for your Buildbotics controller) and it gives you a nice graph and shows you the rpm limits for your motor. Hope this helps Adam
Thanks for all of the great input and recommendations! I did a few experiments to see how I could increase the speed of the stepper. Power Supply I upgraded to a 36V power supply and what a difference it made! I got 1.5X (39m/min from 26m/min) the speed before stalling out the motor (36/24 = 1.5, imagine that). Microstepping I set micro stepping to 32. Going lower or higher than this resulted in the motor stalling earlier. The Open Builds motor seemed more sensitive than the stronger motor described below. Motor For all of the above, I used OpenBuilds High Torque motor. Even with the improvements, performance was a bit lackluster. I tested CNC Router Parts/Avid CNC 3/8" shaft motor and had really amazing results. Using the CNCRP motor, I never stalled it and got max speeds of 130m/min which is well beyond what I'll need. All the testing was with no load on the motor, but the 36V power supply really brought out the best in the motors. Acceleration/Jerk I didn't thoroughly find the maximums of the acceleration and jerk settings but did it by feel for my application. For the OpenBuilds High Torque motor I used max-accel = 1100 km/min^2 and max-jerk = 550 km/min^3. For the CNC Router Parts motor, max-accel = 2500 km/min^2 and max-jerk = 1500 km/min^3. The smooth acceleration of the controller results in super smooth action. Thanks again for the advice.
The inductance of the cnc parts motor is 3mH vs 4 for the openbuilds.. this will get you higher possible rpm as well.. at the cost of higher amps I'm surprised that 1/32 stepping is working best? Gary
The inductance difference makes sense. Yes, I was a bit surprised that 1/32 micro stepping worked the best on the OpenBuilds motor. It seems to be quite a sensitive motor. The Buildbotics controller has a 250k step/min max so it's possible to try high step rates.
