LS Electric L7C & L7P Servo System Auto Tune Part 2 from AutomationDirect

?On-line tuning is similar to the off-line tuning we did in the previous video. The difference is instead of issuing three pulses to do the tuning, it tunes the servo system to whatever waveform is supplied by an external PLC controller, or the servo's internal indexer while the system is running. For this video, we’ll use the servo’s internal indexer to generate the position commands. There’s a whole video showing you how to set that up so I’ll very quickly walk through that setup. Do a factory reset. Wait a few seconds for this dialog to go away - got it. Reset the controller. Wait for it to reconnect. Setup wizard - go online. Yes, we want to read the drive's current configuration. We want indexing mode and we’ll control indexing from the digital inputs so we’ll turn this off. All of this is fine. And for digital inputs we want a servo enable, we’ll use the first three index selectors, and we’ll need a start and a stop. The rest of these aren’t needed. I like to see the index outputs – they are handy for debugging things. And I like to keep Homing Mode 35 handy so I can quickly reset the position register. Write that to the drive. Again, if you want to see how to set all of that up, watch the indexing video. The write was successful. Reset the controller to make sure we get a clean start. That’s really important – if indexing doesn’t work for you, it’s probably because you missed this step. Always reset the controller after making changes to the drive. I've already set up an indexing motion profile. It just says, rotate 100,000 user units at this rate, using these accelerations. Do it once, wait 200 milliseconds, then go to Index 1. Index 1 says go back 100,000 user units using the same velocity and accelerations. Wait 200 milliseconds, then go back to Index 0. So this back and forth motion will just repeat over and over. Let’s bring up the scope which I've already set up with the commanded position – this software calls it the position demand – and the actual position. I also set the vertical scale to show us the 0 to 100,000 user units of motion we set up. If it isn’t already up, click this guy to bring up the digital inputs. We’ll click on these guys to override any wires or actual switches that may be connected to the drive. I love that I can override any wiring that may might be installed. Ok, all we have done so far is set up a motion profile using indexing. Again, we cover all of this in this video on the indexing mode. Now that we have done that we can finally talk about on-line auto-tuning. We just did a factory reset, so this drive has not been tuned yet. So let’s see how it performs out of the box with this load. I’ll start the scope, go to the digital inputs and enable the drive output, make sure we are starting at position index zero – yep these index selectors are all off – and start the move. Sure enough, we see the commanded position in white go out a 100,000 user units, wait 200 milliseconds, then go back a hundred thousand user units, wait 200 milliseconds and repeat. If I stop the scope and zoom in, we also see the red actual position is basically tracking the commanded position, but not very well because the servo system hasn’t been tuned yet. So let’s tune the servo to improve that. I’ll start the scope again so we can watch things change. Zoom messed up our axis so I’ll enable and then disable these to get it back. Click the tuning fork to bring up the tuning dialog. Check this box to enable on-line tuning. As soon as I do that the tuning changes because of how the rigidity is set up. In on-line tuning we can control the rigidity and this adaptation speed. This distance is only for off-line tuning and doesn’t apply here. The adaptation speed is just how quickly the tuning adapts. Our demo is a fairly slow process, so I’ll keep that number small. You would only need to speed this up if you have a really fast process that you need tuning to adapt to really quickly. The key thing though is this rigidity. If we take it real low, the red actual position doesn’t track the commanded position very well at all, does it? As we make it more and more rigid – I’m just clicking on the slide bar to change it one unit at a time - the red actual position looks more and more like the white commanded position. Let’s stop the scope and zoom in again. Now it’s tracking really well with just a short delay between the requested position and the actual. It takes time for the drive to do what it told it to of course. Let’s get the scope running again, and I’ll toggle these to quickly reset the scale. How cool is that? With on-line tuning, you can see the tuning process in real-time and can tweak it simply by moving this one slider. If you want to see the results each time you change the rigidity just hit the read button down here to refresh the display. That also updates the full gain window, this guy - if you prefer to use that. You could also refresh both displays using this. I’ll reduce the rigidity and hit read. We see the gains reduce – exactly what we expect. Reduce it again, the gains reduce even more. Perfect. Let’s get back to rigidity of 10 and refresh that. This is the estimated inertia mismatch, which is fine because this servo system can handle up to a 30 to 1 mismatch! You may be able to tune larger mismatches with some effort, but the system performance isn’t guaranteed beyond 30 to 1. This is important: Depending on your system, it is possible to get too rigid. I’ll keep increasing the rigidity and oops! The drive couldn’t maintain position as rigid as we requested so it gave up and issued an alarm. It says, Nope, can’t do it! This alarm says the deviation between where we wanted to be and where we actually are is just too large and the drive can’t pull it back in. Which is what we expect. So be careful you don’t overdo tuning – you can get it to the point where it becomes unstable. I would even suggest going until it fails like this – if your system allows it of course - and then back off a few rigidity steps so you know you have margin before things go unstable. One caution - there is another software quirk when doing tuning and indexing like this. Normally, you would simply hit this guy to reset the alarm and continue on. And that does reset the alarm, but the motor isn’t moving. In fact, there’s no torque at all! Let’s hit stop and then disable all the indexing controls. If I go to jog, enable the drive, and try to jog the motor, hmmm, jog isn’t moving the motor either. Why not? I’m not exactly sure, but I think this free drive control software didn’t reset everything it was supposed to, so the drive is stuck. The drive thinks it’s working, but there is no power being applied to the motor. There’s some kind of output inhibit that didn’t get reset. The only way I can get this software to fully reset the drive is to re-do everything we did at the start of the video. A full factory reset, give it a second, there it is. We always do a controller reset after a factory reset just to make sure everything is cleared out. Wait for it to re-connect, and then walk through a full setup wizard and reset everything to where it was – I’ll quickly do that – and write that to the drive. And we always do a controller reset when we change drive modes just to make sure we get a clean start. Wait for it to reconnect and we’re good to go. Again, If indexing doesn’t work, it’s probably because you missed this controller reset step. We did this re-configuration step just to clear out the alarm condition because my version of the software and firmware don’t seem to re-enable the motor after clearing the alarm. Hopefully, you won’t have to do that in future versions. But if you do, now you know how. If we look at indexing, we see the factory reset didn’t erase the indexing commands. If I refresh indexing - yep, you can see all the indexing is right where we left it. Let’s get our scope back font and center. Go to digital inputs, enable the drive, make sure we are starting at Index 0. We are - and start indexing. And if we go to tuning and click on on-line tuning we can adjust the rigidity, but this time let’s stay away from the 12, 13, 14 numbers. Looks like around 10 is all we really need. And that’s the key takeaway from all of this. Don’t overdo rigidity. The farther you are away from the rigidity that caused the alarm, the more stable your system will be. If you increase the adaptation speed, it will do a little better but it will still eventually fault out. The bottom line is: You always want to tune your servo to get the best possible performance. And if possible, use the on-line version so you can tune it to your system's position commands. And there’s really no excuse not to tune your system because this free DriveCM software makes it super easy to do. You can download that right now at AutomationDirect.com. Click here to learn more about the LS Electric servo system. Click here to learn about AutomationDirect’s free award-winning support options and click here to subscribe to our YouTube channel and hit that little bell icon so you will be notified when we publish more videos like this one.