3/31/2022 6:54 PM | |
Joined: 9/22/2020 Last visit: 8/19/2022 Posts: 15 Rating:
|
I am running a Tension control application on my machine. The machine has an unwinder a winder an LTC and a leading axis. The web goes from unwind axis to an LTC axis then the leading axis and then the winder. In one mode the web needs to be synchronized with the web of another machine. I am doing this by:
When I run tension control in the gear locked mode, I see a lot of vibration on the web. But without the gear-locked mode, the Tension Control unit seems to be ok. I checked the variable gasWinderDiag.sActLineAxisMotionVector.r64LineSpeed and in gear-lock mode this value seems to be oscillating and my assumption is that this oscillation is causing the vibrations, so my question is:
|
Last edited by: AutoGen_5045471 at: 03/31/2022 22:48:32 |
|
3/31/2022 8:59 PM | |
Joined: 7/7/2010 Last visit: 3/10/2025 Posts: 15721 Rating:
|
Are you using Siemens' Winder and Tension library as part of your system? https://support.industry.siemens.com/cs/document/58565043 You might want to do a search of these forums for that same library to see what Product Conference seems to find the best answers for situations closest to your own. It might be the SINAMICS general conference would be better - if you are using Siemens SINAMICS drives.
|
science guy |
|
This contribution was helpful to
1 thankful Users |
3/31/2022 9:52 PM | |
Joined: 7/7/2010 Last visit: 3/10/2025 Posts: 15721 Rating:
|
Depending on the type of encoder, how much backlash and mechanical "slop" between encoder and physical load (gearing / linkages / motor shaft / gearbox and so on), you may need or want a variety of filtering capability. If the tension control application / library has an "auto-tuning" option, I recommend using that for each operational mode, saving each operational mode to a unique instance-DB (along with appropriate logic - and maybe even a "transition" instance DB for each type of transition) to help make bump-less transitions between modes). If there is no auto-tune capability, are you savvy enough (with available people and online information) to create and integrate that capability yourself? All that said, you may be able to simply insert a simple filter in between the encoder signal and whatever uses the encoder feedback, and adjust the tuning until the vibration is gone while in Tension control (with that encoder). Siemens has a library, LGF, Library of General Functions, which has filter blocks available. If the vibration is detectable via PLC software from an existing external device (like small current, speed, torque, tension, accelerometer, etc), you may be able to make an "adaptive" filter. If so, it could work something like this:
Hopefully that gives you an idea on how to go about manually and/or adaptively tuning. Of course for adaptive tuning, you need some max and min values for filtering to prevent instability or to let people know something is physically wrong - before something breaks. Adding a 30 second filter for example masks the problem to the point where things can run-away without anything slowing it down fast enough. If you are very control theory savvy, ignore most of everything I've said and post what you really do!
|
science guy |
|
4/5/2022 1:41 AM | |
Joined: 9/22/2020 Last visit: 8/19/2022 Posts: 15 Rating:
|
I was looking at the external encoder signals and realised that the time-period between two square waves of the encoder is around 16-20ms. Is this an expected value or is this too coarse for an external encoder? I am guessing this coarse encoder resolution is causing the vibration, is this possible? |
Last edited by: AutoGen_5045471 at: 04/05/2022 01:41:59 |
|
Follow us on