pH Probe/Analyzer issues with VFDs

We installed a pH control system in our wastewater system.
pH probe report to a Rosemont analyzer and send a 4-20 signal to our controller and then to Ignition SCADA.
One of the tanks (Tank 2)has the pH maxed at 22 mA and the first tank with the same wastewater is functioning correctly. When the agitator motors are turned off (They are driven by PF 525s) The pH returns to normal. This only occurs in tank #2. The probe and analyzer has been verified. Bonding and grounding of motors and drives and conduit verified as well. When you place the suspect probe in tank 1 with the drives running, it reads normally. Take it to tank 2, drop it in and it maxes out immediately. Again, turning off the agitators on both vessels the probe reads normal until you enable either or both motors and then it's maxed out.
This probe is brand new and acts exactly as the first one installed.

I suspect it's an issue with the VFDs and have recommended noise filters on the terminal/load side of the drives. Does this make sense?


Thanks

Noise on the drive output (you are referring to t1, t2, t3 or u, v, w, I believe) should not cause problems with the analog signal, unless your shielding is faulty and the wires are in very close proximity. You will want delta V delta t filters on the drive output if the run is long to prevent problems in the motor. I would look closer at the wiring to drive #2. Is the analog to the drive coming from the controller?

Robb
The analog speed reference to the drive is Ethernet and run in different conduit.
Everything is brand new, it’s a new install.
I do not necessarily believe there’s a problem with the analog signal, it’s generated in the Rosemont analyzer module which has a board where the probe is wired into. It takes the RTD wires and pH wires (both enclosed in the probe) and converts to dual 4-20 outputs to PLC input. We can speed up the drive and it takes more time to fault the sensor; if you slow it dow to say 15hz it’ll fault immediately like there’s more noise generated at lower frequency.

How did you run the wiring from the drive to the motor? Shielded VFD cable? Loose conductors in conduit? If conduit is it steel conduit? If steel, is is boded everywhere and all bonding connections have been checked?

The reason for all this questioning is because it has the symptoms of EMI/RFI from the VFD output bleeding into your analog signals. The output cables from a VFD MUST be properly shielded, either with steel conduit that is fully bonded to ground at both ends, or shielded VFD cable with the shields grounded at BOTH ends (unlike instrumentation shielding that is done at only one end).

The contractor ran it loose thhn in PVC conduit.
Not shielded cable and it runs from a job to motor in sealtite.
I suspect it should have been in shielded cable bit I didn't organize this event.

There you go then. Those output cables are acting as radio broadcast antennae. FM in radio stands for Frequency Modulation; and what does a VFD do?

And for the purists, yes I’m aware it’s not exactly the same, I’m just illustrating a point that not paying attention to this issue is asking for trouble.

Would a ferrite choke help on the drive output?

toddp65 said:

The contractor ran it loose thhn in PVC conduit.
Not shielded cable and it runs from a job to motor in sealtite.
I suspect it should have been in shielded cable bit I didn't organize this event.

Click to expand...

I would highly suggest replacing this with screened VSD cable that is properly terminated at both ends, including screen continuity through the local isolator. At the moment what you have is a long noise radiating antenna in the vicinity of a very sensitive instrument that operates in the mV range for the signal from the probe to the transmitter

Even when things are properly done, you can sometimes still experience issues at certain frequencies. I have a deep water bore site with a drive that would send the flowmeter into a tailspin at 33-35 Hz and nowhere else. Had to program the drive with skip frequencies to simply jump over this trouble band

Sounds like you have unshielded analog cables and VFD power output noise. Shield one or both. We use a lot of Teck cable, rarely noise issues like you are seeing, but always use shielded analog wiring, even in conduit.

Thanks all.

The probe cables are shielded and the cables to the controller are as well.

This is heading back to the unshielded 480 wiring to the drives.

The contractor is supposed to install filters on the drives this morning--would this eliminate the need to run shielded motor cables...maybe?

toddp65 said:

Thanks all.

The probe cables are shielded and the cables to the controller are as well.

This is heading back to the unshielded 480 wiring to the drives.

The contractor is supposed to install filters on the drives this morning--would this eliminate the need to run shielded motor cables...maybe?

Click to expand...

Are the probe cable shields grounded at exactly one point. Not grounding the shield, or grounding both ends can do more harm than good.

Also, is it possible to move probe #1 into tank #2? That may re-focus troubleshooting on the instrument if probe #1 does not exhibit the same problem.

The probe cables run to a terminal board in the analyzer. there are four wires; two are RTD for temperature the other two for the pH. The wire is shielded at the connection point but there's no way to determine on the probe itself. Now as for the 4-20 from the analyzer to the PLC, the shield is grounded in the PLC panel.

Using an output filter without screened motor cable could be enough.
Also, using an output filter could be a better solution than screening the cable, as a filter also protects the motor windings and motor bearings.
Output filter AND screened cables are even better, but it may not be necessary depending on the installation.
And there are different grades to the ouput filter. Next step could be a du/dt filter, which may be a solution if the problem persists, and replacing the cable is too difficult or too expensive.
How long is the cable between VFD and motor ?

Mispeld said:

Are the probe cable shields grounded at exactly one point. Not grounding the shield, or grounding both ends can do more harm than good.

Also, is it possible to move probe #1 into tank #2? That may re-focus troubleshooting on the instrument if probe #1 does not exhibit the same problem.

Click to expand...

The odd thing is, when we put the probe into tank 1, there's no issue--it reads as designed with all drives running.
When you move it back to tank 2, it faults out immediately if one or both of the drives are running.

These are fiberglass, double walled vessels. The waste water from tank 1 flows into tank 2 via PVC piping.

JesperMP said:

Using an output filter without screened motor cable could be enough.
Also, using an output filter could be a better solution than screening the cable, as a filter also protects the motor windings and motor bearings.
Output filter AND screened cables are even better, but it may not be necessary depending on the installation.
And there are different grades to the ouput filter. Next step could be a du/dt filter, which may be a solution if the problem persists, and replacing the cable is too difficult or too expensive.
How long is the cable between VFD and motor ?

Click to expand...

It's only about 20 feet or so (7 meters). They have the wiring in PVC conduit then to a fiberglass J/B with no disconnect wire nut to motor (???).
They apparently have ordered the filters and will run cable within the next day or so.