PID tuning of GALVOS

I am developing a laser scan head for a laser show using BLDC motors. I am facing a problem with regards to the tuning of the motors.

I have obtained the transfer function for the BLDC motor and plotted its open loop step response. Using the Zeigler NIchols open loop method i have calculated the coefficients - Kp = 5, ki=1800, Kd=0.004.

However on feeding this values to the BLDC SERVO Drive the mirrors start oscillating. Why is this so?

Moreover on using the autotune software, kp=155, ki=7,kd=1000 are the values obtained. The system does not oscillate and works fine with this order of values. Why is there such a huge difference in the values obtained and which one is correct? what are kp ki kd magnitude order generally in use?

In your calculations, do you have all the units correct and do you have the loop time factored correctly? Your Integral looks ridiculously large.

http://www.photonlexicon.com/forums/...cons/icon1.png

Moreover on using the autotune software, kp=155, ki=7,kd=1000 are the values obtained. The system does not oscillate and works fine with this order of values. Why is there such a huge difference in the values obtained and which one is correct? what are kp ki kd magnitude order generally in use?[/QUOTE]


A galvo is NOT a BLDC in structure.

Your going to have to ferret this out for yourself. No one short of the factory looks at digital Galvo amplifiers in terms of Zeiger-Nichols. Only perhaps 10-20 people in the world would know that, and only one of them posts here, and he is strongly disinclined to help you for commercial reasons. (Bill B.) What is the shaft inertia ? It is on the order of 0.018 to .1 gm/cm^2 for a graphics scanner. The mirror inertia is no more then 1/10 to 10 times the shaft inertia.

http://www.camtech.com/products/6800/6810p.html

http://www.camtech.com/products/6200/6210.html

The book is:
Laser Beam Scanning
Gerald F. Marshall Editor
Low Inertia Galvanometer Scanners
Start at around page 247

Steve