I am using your A4983 with voltage regulator to drive NEMA23 stepper motor from Lin Engineering. The motor coil voltage and current are 4.2V and 2.12A respectively. The power supply came from 12V battery. I use PIC18F6680 to control the inputs.
Every time when I connect the step input with PWM on the board, the motor controller is disable (there was no current through the motor coil). And when I disconnect the PWM on the step pin, it is normal.
This is how I wired the board.
GND - connect to MS1,2,3
5V - join with Vdd by SMT pad
Vdd - join with 5V by SMT pad , also connect to RESET and SLEEP
3.3V - no connection
GND - no connection
REF - 0.55V
Enable - Logic low from PIC RC0
MS1 , MS2 , MS3 connected to GND
RESET , SLEEP connected to Vdd
STEP - RC2 PWM from PIC
DIR - Logic output from PIC RC1
VMOT - 12 V battery positive , also supply to PIC board
GND - 12 V battery negative, also supply to PIC board
2B, 2A - to motor coil 1
1B, 1A - to motor coil 2
Vdd- no connection
GND- no connection
I have tried with a different PWM frequency 156.5kHz, 312.5kHz and 415.6kHz with duty cycle 50%, but nothing was different. The oscilloscope showed that the PWM was correct. I have measured the voltage across each input, there are all correct.
Anything I could make it wrong??? I am really stuck.
It sounds like you could debug this problem with some relatively simple testing. If the driver behaves normally when the step pin is disconnected and fails when it is PWMed, break things down into the two sub-cases:
What happens when you hold the step pin high?
What happens when you hold the step pin low?
If one of those two cases leads to strange behavior, you have narrowed down the problem to something that can be explored further (and without complicating things with a PWM signal). If neither of those cases leads to strange behavior, then you have determined that the problem is with the PWM itself and you can explore it further by seeing at what frequency things start to fail (I suggest you start testing at a frequency much lower than 100 kHz).
It works at 2.44kHz with 1/16 step resolution. Thanks Ben for the help. I still don’t understand the principle around PWM frequency, step resolution and current limiting. I guess they are all related.
Every pulse to the step pin rotates the magnetic field in the stepper motor by one step (or one micro step), and you are counting on the rotor to follow and realign with the field before you take the next step. The rotor’s inertia limits how quickly it can respond to the changing field, so if you rotate the field faster than this inertial limitation, the rotor simply won’t move (or it might vibrate in place). Does this make sense?
You will find that your maximum step rate is limited by the amount of torque you need.
Yes, it makes sense, thanks for the explanation Ben. I actually was looking for some source of equations to start with, but it seems too complicated. It is better to start with a low step rate and then increase it later. By the way, I got the linear actuator working and ready to drive a robotic arm gripper.
The trial-and-error approach is probably good enough, but you can be a bit more rigorous if you have enough data about your stepper motor. The documentation for many stepper motors includes a pull-out-torque curve (see our stepper motor product page for an example) that shows the pull-out torque (the torque beyond which the rotor will not track with the rotating field) as a function of step rate. If you can come up with an estimate for the torque on your stepper motor, you can use this curve to determine your maximum step rate. Note that this curve also depends on the current you are supplying the stepper motor and the driver voltage (the higher the voltage, the quicker the current ramps up, which in turn allows for faster step rates or higher torque). Also, you can often reach higher rotation speeds if you gradually ramp up your step rate rather than starting at your desired step rate from rest.
I’m glad to hear you have your actuator working. If you have any pictures or video of your gripper mechanism in action, we’d love to see them!
