Motor never leaves to oscilate

[luis_gts]

Hello,

Recently, I bought 3 viper 200 drivers and 3 renco encoders.
I'm really confused about how to tunning the PID. I already read the manual
and follow the steps one by one.

Scenario:

I'm working in an Anayak 8 Anayakmatic retrofit, I'm using the original motors
I installed the encoders in the screw which moves the mill's table. Those screws are not directly connected to the motor, the transmission is through a band. V2.2 is the viper driver version.

I connected the viper drive with 12 volts, also, the rs232 interface and motor power supply, encoder and motor, and finally the step and direction from the break out board. I'm using Mach3

Problem:

I started with the X axis, without decoupled the motor from the transmision which moves the mill's table, to have the encoder working

I started the hyperterminal and configured p with 400 and d with 800 (I saw a recommendation in this forum)

In one side of the screw, which moves the table, it's the encoder and in the other side is posible to see it, for the oscilation.

in the movement of the table through X axis, there is no oscilation in the table, but in the moment I powered the motor, in the screw I can see an oscilation.

I increased kp, the oscillation diminished aprox. 6000, then increased kd aprox 10000, but the oscilation never stopped. (oscillation in the screw)

well, sorry for all this explanation, but I considered it was necessary.

Questions

1.- Because I'm really new working with pid, I would like to know if the screw, that oscillates, should be totally stiffed (without any movement), or there is a safe rate.

2.- Which are the maximum kp and kd values to configure PID, because I started with 400 and I finished with 6000. so I want to be sure that 6000 is an allowed value, that couldn't affect the viper performance or just too high.


Thank you

[H.O]

Hi Luis,
I have no personal experience with the Vipers but in this scenario, with the encoder on the screw you need to make absolutely sure that there is NO backlash between the motor and screw. In other words, the belt needs to be very tight.

If the screw is thin and long there's also the possibility of the motor "twisting" the screw slightly before it begins to turn, this will also look like backlash from the motors point of view and if this is what is happening the oscillations would be different depending on where on the screw the nut is at the moment - I think.

[Larken]

Increasing the P and D too high will increase oscillation. over 3000-4000 is very high in settings.
You didn't mention what the L (loop) parameter is but this is a important parameter that needs to be set at the start of tuning. L would be between 3 and 6 for your size motor.

Try P800 D1600 and adjust L for most stable. Then increase D to no more than 3000.
How stiff the motor should be, depends on encoder resolution and whats needed. Is your motor a 4 pole or 2 pole ? A 4 pole will be stiffer. All motors will oscillate if gains are too high.
Don't get carried away. Tune it moderate, run it, and then see if you need it tighter.

What is the Powersupply voltage ? What is the motor rated voltage ?

Like HO says though, you can't have any play between motor and encoder or its very hard to tune. But there is a hysterysis command (h) in the vewer versions that can create a dead band for overcoming this.

The firmware version 2.2 is a few years old. Most of the improvements since are for current control and drive protection, but PID tuning hasn't changed.

Are your drives ones with a socketed cpu chip, of are they surface mount ?

Larry K