New to animatics, i am using 4x SM2337DT-PLS2(firmware 4.95) for an xyz gantry. all servo controlled by a controller with step/dir mode. controller send step signal with 1000000 khz and signal width 2 uS.
I observe an inconsistency between pulse count in SMI from same distance move command from external controller. which lead to step loss. I make the signal width upto 7 uS with little success.
below is the program i am using in all motor.
UAI UBI UDI UCI ZS MS MFMUL=4 MFDIV=7 MSR G END
My question is:
1. What is the minimum pulse duration/width to recognize one pulse by servo?
2. can i use Sine communication mode(MDS) for smooth movement.( i noticed jerky motion) , if not what are the alternative suggestion?
Thanks in advance.
Hello Joy,
When you say you observe an inconsistency in SMI, through what means are you comparing the pulse? Sampling in SMI is limited to the specs of the computer you're running it on.
As far as MDS goes, because you're using a Class 4 Legacy motor, sine commutation mode is not available. This is an option in the Class 5 firmware.
I am giving 100mm move command through step/direction to servo and watching step pulse count in SMI software connected to servo by serial cable.
I can see count for same 100mm command shows sometime 7900 pulse count, sometime 8200. My understanding was pulse count captured in servo which showing in SMI software tool?
I am using SMI Tools>motorview to monitor the count of pulse. I didn't know its limited by computer. is it true that pulse count showing in motor view is limited to specs of the computer?
Yes, it is true that the accuracy of sampling is to an extent limited within SMI. This is in part due to the other simultaneous processes within your computer. You can track the CLK value as well to get a more accurate reading of time. What is your goal in commanding and monitoring the motor in the way that you are?
I am using 4x SM2337DT-PLS2 for a 3D printer/CNC. I am using smoothieboard to calculate path and send it through step/directory to servo. I found it would be the easiest solution as I believe it would still out perform a stepper and it's close loop.
After i set up everything, I noticed my print is shifting towards one direction. That's when I start to investigate pulse accuracy and duration.
Is there a sample program what I can upload to capture pulse count to rule out pulse duration from the issue I am having?
The pulse width will be limited by the max frequency, which is on the order of MHz. If you are noticing drift, I would check your wiring first, slow down the process, and monitor for noise at low speeds.