Follow Parker Hannifin on social media:
I have a pair of MX80L with the 10 nm encoders that I am trying to set up with some vix-ah drives. When I try to get the servo loops tuned I see serious cogging in the open loop moves where the allowed input does not seem to be able to move past the valleys sometimes. If I increase the input level when it accelerates into a valley it trips the encoder overflow fault. Here is the 1status serial response:
Most likely, the root cause of this issue is a commutation problem. The encoder, hall sensors and motor phases all need to line up in a very specific way to ensure that the drive can deliver current to the correct phase and generate controlled motion.
When halls, phases and the encoder do not align correctly, you can sometimes get the motor to work a little bit. Since the drive is delivering current at the wrong electrical angle, the motor can lose force in certain areas, which looks a lot like bad cogging. Most of the time, the drive will also generate thermal faults if operated too long under these conditions.
Since the bad motion leads to a lot of jumping around, high resolution encoders can also suffer from overflow faults -- basically just being driven too fast.
It looks like your encoder resolution is set correctly from the image above, so there are two likely causes for this problem:
There are also a couple of less likely possibilities:
I hope this helps. Please let us know if your issue persists.
Thanks for the reply. The colors of the connections match the labels in the two manuals/ I have the encoder interpolator boxes installed. After the faults I usually 1Z them with the serial connection.
I worked on them again this morning and got them both to work but the CL levels have to be tuned very low down like 5% and 7% which gives them very narrow margin for dealing with any issues. Touching the stages will generally fault them as the servo loop is not strong enough to prevent motion and if the Loop is off then the speed is often faster than the counter allowed rate.
What controller are you using with this ViX? Is it a Parker controller or third-party? It could be that the ViX and MX80L are fine and that the controller just needs to have its gains adjusted.
It is am Omron controller. I think I have this sorted mainly by restricting the CL in both the amplifier and then the max output in the omron close loop to make sure it does not ever trip the encoder fault on the Vix. That at least gets the motion smooth but the servo loop can only hold back about a pound of force so the loop will crash anytime the user touches the platform likely. When the loop is killed the platform can be touches but it is pretty easy to generate the 30 mm/s instantously velocity just touching the stage. This also faults the system. These faults need a power reset to the drive since their is no control pin connection which provides the 1Z from the terminal equivalent.
In general, the MX80L is not designed to apply much force, only 0.9-1.8 lb continuous depending on the model. My general recommendation is that the machine needs to be built such that the user does not move the stage. In addition, the travel path should be free of obstructions. If you absolutely need a reset feature equivalent to 1Z, you could use a relay to cut power to terminal X1/7 (+24 VDC) and reapply it a couple seconds later.
In general, the MX80L is not designed to apply much force, only 0.9-1.8 lb continuous depending on the model. My general recommendation is that the machine needs to be built such that the user does not move the stage. In addition, the travel path should be free of obstructions. If you absolutely need a reset feature equivalent to 1Z, you could use a relay to cut power to terminal X1/7 (+24 VDC) and reapply it a couple seconds later.
