Want to add something to this that bit us hard: test your backups.
We had a beautiful backup routine — quarterly backups, three copies, everything labeled and dated. Then a Yaskawa drive failed and we went to restore from the .ysp file. File was corrupted. Tried the second copy — same file, same corruption (it was copied from the first). The printed hardcopy was from 6 months earlier, before a major re-tune.
Now our process includes a verification step: after every backup, we do a compare. On Yaskawa, SigmaWin+ has a parameter compare function — load the backup file and compare against the live drive. Every parameter should match. If even one doesn't, the backup is suspect.
For Siemens S120, we learned another lesson: the memory card backup includes the firmware version. If you replace a failed drive with a newer firmware revision (which happens — you can't always get the exact same firmware), the parameter file may not load cleanly. Some parameters get renamed or renumbered between firmware versions. We now keep a text export of all parameters alongside the binary backup. The text file is human-readable and can be manually re-entered on any firmware version.
Fanuc-specific gotcha: the FASSB backup saves servo parameters but NOT the PMC (ladder) parameters. If your press has custom PMC logic for die protection, tonnage monitoring, or recipe management, that's a separate backup through the PMC menu. We lost our custom die protection logic once because we thought the servo backup covered everything.
Our current backup schedule:
- After any parameter change: immediate backup + compare
- Monthly: full backup of all drives, PLC, HMI recipes
- Quarterly: test restore on a bench drive (we keep one spare drive per brand specifically for this)
- Annually: full system image including PLC program, HMI configuration, network settings, and drive parameters
The bench drive for test restores costs about $2-3K per brand. Sounds expensive until you compare it to 3 days of downtime because your backup was bad.