MAX - Reducing Encoder Resolution

[rustawilliams]

Hello, I was wondering if there is a way to reduce the number of counts read from my encoder per unit rotation (currently at 1000 counts per revolution) by the MAX. Currently, it is apparent that I need less than 1 proportional gain to get a good shape for further calibration and this cannot be done with MAX. With a lower resolution, this problem would be solved as I would be able to do a large enough unit step for my step response test so that proportional gain can be changed to a suitable value (an integer).

Your help would be much appreciated. 

[BCho]

Hey rustawilliams,

The encoder resolution is more a function of the encoder itself and less a function of MAX. The resolution of the encoder dictates the number of pulses per revolution of the encoder. So, in your case if your motor moves one rotation, your encoder will give 1000 digital pulses. The resolution you see in MAX tells NI Motion how many pulses from the encoder to expect per revolution of the motor. If you were to set this resolution in MAX to say 2000 and tell your motor to move 1 revolution, your motor will only actually move 2 rotations. So, you can change this setting in MAX, but it won't change the actual number of pulses being sent from your encoder.

[Jochen]

As Ben has already said, reducing the physical resolution of the encoder is not possible with an NI-Motion board (by the way, which board are you using?). Still there might be other options to solve the issue:

1. The default PID cycle time of NI-Motion boards is 250 µs. If you are using a 7340 board, each activated axis requires a time slice of 62.5 µs. So if you are using e. g. only 2 axes, you can reduce the PID cycle time to 125 µs, which will give you some more headroom for setting the PID parameters (less encoder counts per PID cycle).
   
2. If your system becomes unstable with PID settings of Kp=1, Kd=0 and Ki=0, I doubt, that the encoder resolution is the root cause of the issue. In this case it's much more likely, that something else is wrong. The first candidate to look at is the encoder polarity. Please change the polarity of one of the two phases (either A or B). Is the system still unstable?

Kind regards,

Jochen Klier

National Instruments