Near miss on the cartoner last night and I want to understand it before I touch any code. An operator was jogging the infeed servo from the PanelView Plus 7 when someone pulled the Ethernet cable at the panel to plug a laptop in. The axis kept jogging for about eight seconds and ran into the end of travel switch, which stopped it. Nobody hurt.
CompactLogix 5069-L320ER, Kinetix 5500 on the infeed, Studio 5000 v33. The jog is a momentary push button on the HMI writing a BOOL, and the PLC jogs while that BOOL is 1.
I checked the button, Set on press, Reset on release, then pulled the cable myself on a test run and it just carried on jogging. Nothing has changed on this line, the code is two years old.
Is this the button hold time? Could someone explain what's meant to clear that bit?
What's the jog velocity, and is there a hold time set on that button? Some people set a hold time of a few seconds so the press registers, and that'd keep the bit high after release. Also which firmware on the PVP7 and what's the coarse update period on the motion group, just to have it.
Hold time is 0. Jog velocity is 200 mm removed link PVP7 on v12, coarse update 4 ms. So pretty sure it isn't the button being slow.
Pulled the cable three more times watching the tag online. The BOOL stays at 1 until I plug back in and press the button again. It never drops by itself.
Small point first: a jog from an HMI is a standard interlock, not a safety function. If the risk assessment needs the jog to stop reliably, that's a hardwired hold-to-run or a safety-rated input, not a screen button, whatever you fix in the HMI. That's true and also not your bug. Your bug is that nothing in the PLC owns that bit once the terminal is gone.
A momentary button is two writes: 1 on press, 0 on release. The release write never arrived because the cable was out, and nothing in the PLC clears the bit, so it stays 1 forever.
Fix on the PLC side. Global Connections, Date and Time tab, point the Second output at a DINT in the controller. The terminal writes it about once a second. In the PLC, a TON that resets when that value changes, 3 s preset. When it times out, clear every HMI command bit and drop manual mode.
There's a writeup on HMI to PLC handshakes here: https://plctr.com/introduction-to-hmi-human-machine-interface-in-plc-systems/
Added the Second output and the TON. Pulled the cable again. Jog stopped at 3.2 s. Better, but during the test the axis also stopped once with the cable in, for about 400 ms, then carried on. Not sure what that was.
Sounds like the watchdog tripping on a late Second update. I reckon 3 s is tight for a terminal that's also drawing screens. Compare against the last value with a one shot and give the TON 5 s. Trend the DINT and the TON.ACC for a shift and set it from the worst gap, not the typical one.
Preset moved to 5 s, trended a full shift, no false trips, longest gap on the Second tag was 1.8 s. Cable pull stops the jog at about 5.2 s, and manual mode drops so the operator has to re-select it.
So the release never reached the PLC and nothing cleared the bit. The fix was the Second output in Global Connections as a heartbeat and a 5 s TON in the PLC that clears every HMI command bit when it runs out.
Still don't understand the 400 ms stall with the cable in. Longest gap on the trend was 1.8 s, so it shouldn't have tripped. The hardwired hold-to-run is going into the risk assessment review, garyb was right about that.
Thanks hannah, mattr.
Good outcome. Get that hold-to-run in writing before someone forgets.