pH Analog Grounding Problems

Has anyone any experience with grounding problems with pH loops.

I have problems with 2 similar loops comprising of:

Prominent Combination Glass Electrode with integral 4-20mA loop powered transmitter - screened 1pair unarmoured field cable (50m) - Prominent panel mounted pH Controller. Also in series with the transmitter is a 4-20/4-20mA signal isolator (230V ac). The output of the signal isolator goes directly into an anolgue input channel on a Mitsubishi PLC. The 4-20mA output from the pH controller goes to a remote variable speed drive (100m).

My problem lies with the PLC input and associated SCADA/HMI/Telemetry displays. I am able to calibrate the pH system no problem. The controller display is stable and correct. On my Fluke DMM the 4-20mA loops are stable and correct. However the PLC indications are very erratic - swinging by +/- 3pH. If I take the pH probe out of the process and put it in a buffer solution in a plastic beaker the indications become stable.

I spent many hours last night trying to narrow down the cause of the problem - adding/removing screens/earths etc. with little joy. Apart from these two loops all the other channels are ok. I have tried putting these signals onto another channel but have the same problem.

Any advice will be gratefully received

I am not familar with your type of PH probe. Can't you just enter some filtering to the output?

I have had some fits with Miti PLC before, maybe a simple cap across the input would settle it down. You could also use the plc to do a running average.

Hope this helps

Steve

We have similar problems with capacitance level controls in leachate. First take a multimeter and ground the black lead and try measuring the voltage differential both AC and DC of the process.

If there is a differential you might try grounding the solution if that is possible. If not have you tryed the probe in any other position in the process? Higher, lower, left, right?

Do your homework before trying to ground the solution. There could be a shock hazard.

With any analog signals proper shielding to prevent noise induction is a very critical necessity. Make should that the wiring for the analog input is shielded properly (only at one end, prefferably the PLC end) and that the ground bus in the PLC cabinet is properly grounded either to building steel, grounding grid, or a copper grounding rod.

It is also always a good idea to add some filtering to an analog input. This is usually either done by changing a module configuration parameter, or throught the creatation of code.

I have also seen cases where the ph probe was installed to close to upstream metal piping which seemed to cause problems. Same case, when the probe was pulled out of the pipe and placed in a bucket of liquid the reading at the PLC input was stable.

Good luck.