Analog feedback to control motion

[wayne g]

I am trying to control a servo motor using an analog feedback connected to analog input 1 on the PCI 7344 board. The system works ok, except I have no control over the velocity (does not respond to the vel input vi) as well as there is some hysteresis in the system that provides a large deadband area in the operation about the point where the direction should change.

Look forward to your help

thanks wayne

[Jochen]

What exactly does your analog feedback signal represent? The motion board interprets the voltage provided to it's analog inputs as a position feedback signal. Is your analog signal proportional to the motor position? If not, your trajectory could be influenced in the way you describe.
Please take into account that the resolution of the analog inputs of the 7344 is quite low (about +-2048 counts)and thus depending on your motor speed it could become quite difficult for the board to control the speed (especially at low velocities).To improve this behavior you could try to increase the control loop update period and increase the time constant of the velocity filter (flex_load_vel_tc_rs()).
Hysteresis is a problem that can't be eliminated with a PID control algorithm. You n
eed to eliminate the cause of the hysteresis (e. g. mechanical slackness) if possible.

Jochen Klier
Applications Engineering Group Leader
National Instruments Germany GmbH

[wayne g]

Jochen

Thanks for the prompt response.

My analog input signal at present is simply a potentiometer on the bench. This though will eventually be a load cell or any other transducer. The range of this feedback is 0-5V. This is not where the problem seems to lie. The system drives as expected apart from the fact that

a. The driving is all at maximum motor amp output voltage. ie there is no way to say limit the DAC output voltage to maybe 20% of full range and then have this 20% spread proportionately across the complete 16bit DAC output range. This is really one part of my question. This then would enable me to limit the maximum speed of the motor whilst still providing analog feedback control such as if the system is very close to the target then
the velocity is very small and as we get further away the velocity gets larger. ie the system tries as hard as it can to maintain the desired feedback value.

b. The system has a big deadband, not related to any mechanical losses. If for instance I am driving to 1000bits on my ADC. When I reach 930bits the motor is stationery and likewise when I reach 1070bits same. Outside of this range the motor velocity is altered depending on how far from the target of (in this case) 1000bits we are. Then though we get back to the problem listed in part A with the max velocity.

wayne