VFD Autotuning

The very first instance of me using a VFD for hoisting was in 1997. I had used AB's 1336 Impact for the hoists. I had the motor decoupled from the gearbox and performed the rotational tuning. The hoist was rated for 75 Tons load. A load of 50 Tons was our test lift. The hoist lifted it well from the ground. once the hoist was stopped, despite having torque proving for the brakes, the motor would just not lift the load.



On reading the manual, it was then decided to run an Inertia Tuning with the load suspended mid air.



Once Inertia Tuning was done, we had no problem lifting even 90 Tons.


I have always carried out the same procedure with ABB, Yaskawa, Meidensha, and Delta drives. Haven't had any problems so far.


But, reading the Inertia Tuning section in Yaskawa's A1000 manual, got me questioning the validity of my previous projects.
 
Auto tuning should calculate two gains for vertical loads. This is due to gravity. It is also possible to use one gain and a bias. If the inertia or load doesn't change then the gains that depend on time constants shouldn't need to change when using a bias.


On reading the manual, it was then decided to run an Inertia Tuning with the load suspended mid air.
Yes, but there needs to be some way of compensating for gravity. It the controller is good it will use the torque necessary to hold the load stationary as a torque bias. That is what I do. It is also possible the integrator simply winds up to hold the load but that might not work so well when starting and the integrator hasn't found the bias torque yet. In slow applications it may not matter.


When auto tuning the excitation is important. We provide the user with limit on speed distance and rates. There is no need to use square waves like many drives do when the frequency of the motor and load is low.
 
I think there is some confusion here because of the terminology used, some of which probably got inserted by a marketing department. ABB calls the initial routine an "ID Run", which determines electrical characteristics of the motor. Others call this an "autotune". Once the initial routine is done with no load(or light load), then a true autotune can take place under load if the drive is truly capable of this. This will likely only be a speed loop tune, since the torque loop should have been done in the first routine. Most VFD's are not capable of position control, so you will not find the kind of autotune settings(or limits) that you would see on a servo controller or drive.
 
https://deltamotion.com/peter/Videos/RotaryShear.mp4
This video shows a Reliance 250 HP VFD being used for a rotary shear. The Reliance motor can be seen in the lower left had part of the video. The cuts had to be within 0.01. It was cutting sheet metal. We tried to auto tune the motor in velocity mode. It didn't work well because the velocity loop didn't know thatgravity was opposing the motion when going up and was aiding the motion when going down. Basically, the drive controls was only a simple velocity loop controlling the torque without any knowledge of gravity. I suggested that we disabled the drive controls velocity loop and that the control output is now proportional to the required torque. This allowed us to add or subtract torque as a function of where the shear was in it rotation. The motion controller knows where the shear is in the rotation. This compensated for gravity. This was similar to a position or angle feed forward. Then we did an auto tune in torque mode. The controller was able to compute the open loop gain and time constant of the motor and shear. The results were good. You can see the smile on the customer's face at the end. This was done long ago.


Most VFD's are not capable of position control, so you will not find the kind of autotune settings(or limits) that you would see on a servo controller or drive.

I agree that VFDs are not usually good for position control if trying to control a position against a load but smart use of all available information can provide good results. I prefer controlling the torque of a motor when possible as the drive controllers are often stupid and inadequate.


In the office I test with motors in torque mode most of the time.
https://deltamotion.com/peter/Videos/AutoTuneTest2.mp4
What I was actually testing was the picture in picture capability of my screen capture software but I decided to keep this video because the video shows auto tuning a motor in torque mode which is much different from velocity mode or what we do when controlling hydraulics. The key is being able to estimate accelerations. We must do this for hydraulics but it is also handy for controlling motors in torque mode because the torque is proportional to the angular acceleration. Our auto tuning software allows us to set position limits if necessary. If the limits were shorter then the time between the positive and negative pulse would have been closer together so there would have been fewer rotations.


There are auto tuning algorithms and then there are real auto tuning algorithms. Often times companies say they have crude form of auto tuning just to say they have auto tuning.
 

Similar Topics

Hello, I'm trying to delve a little into rs-485 communications for a couple projects of mine. Until now I've been using a delta vfd and a delta...
Replies
2
Views
46
I have an issue with Power Flex 525 during running processing, the VFD stopped suddenly while the PLC and VFD connection ok, VFD does not have any...
Replies
1
Views
87
Has anyone setup communications with Red Lion 3.0 MODBUS to Baker Hughes Centrilift GCS VFD? Thanks
Replies
0
Views
70
Hello, I've been trying to learn this a while now and still have not found out how this works. I have an Omron CJ2M PLC and an ABB ACS 355 VFD...
Replies
1
Views
190
Looking for some assistance. I am very familiar with Yaskawa VFDs, but not so much with AB VFDs. I am working on some hardwired AB PowerFlex 700...
Replies
2
Views
87
Back
Top Bottom