Well, this one was an interesting little puzzle. It's fixed now, but I thought I'd make note of it in case anyone else runs into this.
Recently I was tasked with making a major update to the cell hardware surrounding a KRC4 that is using X11 safety (no SafeOp). The robot had previously been operated simply using an X11 standalone jumper plug, but the customer wanted to add perimeter fences, gates, external deadman switches, etc.
So, I had installed all the new hardware, debugged all the wiring, and worked my way down to where the robot only had one error message left: The generic "KSS00400 Safety Stop" message. No details. By experimenting, I found that the error only affected T1 and T2 -- putting the robot in AUT or EXT modes cleared the error and allowed the robot to run. Which pointed at the External Enabling devices I had just added to the X11 wiring, but I could find nothing wrong with this new wiring -- I tested each signal end-to-end with a continuity checker, and everything was fine.
Even taking a KRCDiag and sending it to KUKA tech support was confusing -- the detailed error messages were pointing to a bad enabler on the SmartPad. But swapping SmartPads between KRC4s didn't have any effect on the issue.
The root cause ended up being one of those stupid things -- the KRC4 has two sets of External Enable signals on the X11. I had only installed one new enabling device, and that was the circuit I was checking. As it happens, there was supposed to be a set of jumpers bridging the unused 2nd channel... they were part of the electrical drawing, but had been left out of the actual build. As soon as I stopped trying to find issues with the enabling device I had installed, and looked the circuit I wasn't using, I found the problem. Adding a pair of jumpers across the 2nd External Enable signals resolved everything.
So, bottom line? "Safety Stop" by itself could refer to any of the Enabling devices -- SmartPad, External 1, or External 2. Gotta check 'em all.
The very vague error message didn't help, though, especially the way the KRCDiag was indicating an issue with the SmartPad (I'm guessing that the SmartPad enable and the External Enables are treated as a single series circuit, in effect). Now, on KRC2 ed2005 controllers, it was possible to get more detailed, individual-signal details from the ESC board. But I haven't found anything equivalent on the KRC4. Is anyone aware of way to get better detail from the KRC4 for this type of error? And if anyone from Augsburg is reading this, that would make a great feature to add.