Lift station. Wet well level is a submersible pressure transmitter, 4-20 mA, into a MicroLogix 1400 with a 1762 IF4. Pump 2 got a new PowerFlex 755 last month, 25hp. Since then the level jumps about 5% whenever pump 2 runs and it goes quiet again with pump 2 off. Pump 1 is on a soft starter and has never bothered it.
I swapped the transmitter for the spare first and that changed nothing, so I put a 2 s filter in the ladder, which hides most of it, except now the pump starts hunt because the level lags.
The signal pair runs in the same duct as the pump 2 motor leads I think. Should I just filter harder in the PLC, or is that papering over it? Is getting that pair out of the duct away from the pump 2 leads worth a shutdown, or is there something cheaper I should try first?
Could be the duct. A 755 puts a lot of edge on those leads. Where's the shield grounded, one end or both?
Both ends. Panel ground bar at this end, and at the junction box on the well there's a ground lug and somebody tied the shield to it.
Both ends grounded turns the shield into a conductor between two grounds that are not at the same potential when a 755 is switching. Current flows on the shield, couples into the pair, that is your 5 percent. Two fixes, shield first. Lift the shield at the well end, tape it back, ground at the panel only. Then routing: signal pair out of the duct with the motor leads, its own conduit, 30 cm apart at least. The PLC filter is a last line, not the fix.
Prove it: pump 2 running, meter in series in the loop, watch the mA.
The shield at one end rule and the input side is in here: https://plctr.com/understanding-the-basics-of-input-output-modules-in-automation/
Lifted the shield at the well end. Jump is down to about 1.5% now. Better, not clean, and I'm not sure what's left. Motor leads and signal pair still share the duct for the last 20 ft into the panel, the rest is separate trays.
That last 20 ft is the rest of it.
Pulled the signal pair out of the duct and ran it in its own conduit for the last 20 ft, shield still grounded at the panel only. Flat now, 0.2%, with pump 2 running.
Two things adding up, then. The shield tied down at both ends gave the drive's noise a path through my signal pair, and the shared duct handed it the rest. One end grounded and its own conduit and it's quiet.
The 2 s filter is still in the ladder, pulling it next week when I'm sure. And I never understood why pump 1 on the soft starter never touched it. Thanks marek, vinnie86.
A soft starter does not switch at 4 kHz. Pull the filter.