Salinity Probe Drift

Hi,
I have been running my Profilux 4 since Jan. this year.
Everything works OK except the following.
My slainity probe (conductivity probe) readings drift so fast in thease days.
It was almost OK until couple of weeks ago.
It suddenly drifts downward from 34.0 psu to 31.2 psu within a week.
I measured actual salinity with a Veegee refractometer and Hanna salinity (conductivity) instrument just calibrated with reliable reference fluid.
The results are very stable within +/-0.1 psu.
I cleaned it with vinegar/rodi 50/50 for 30 minutes, and recalibrated. But it is same.
Is there any other remedy measure to take it back to normal?
Or do I have to replace the probe?
Thanks in advance.

Son

Hello Son,

Unsteady values can be caused from electrical interference and microbubbles forming on the sensor tip.
If you have any power cables running alongside the probe cable, it may cause deviations in displayed values.

Have you dipped the probe in the 50mS calibration fluid? What was the result after a 5 minutes?
Does this "drift" happen at random parts of the day or gradually over the course of hours and days?

Hello Vinny,
Thanks for your comment.
The readings are gradually going down about 0.1 psu down in a day or two, no fluctuation. I had some electrical noise interference on the 1st pH probe in the begining of setup. I rerouted all the sensor cables and segregated from power cables. Also I installed a fed-thru noise suppression ferite core on each probe cable. The problem on the pH probe was cleared after that. I fully understand the interference problem as my background is industrial instrumentation and control system. The phenomenon is not due to the noise. but it seems like due to some contamination or degrade of probe itself, like somthing builds up on the electrodes. Sometimes I stired the probe in tank water, it did not change seriously.
Early this morning, the reading was 29.6 psu. I cleaned the probe tip with rodi water stiring about 5 seconds. Then wiped it with a tissue paper and dipped in 50mS reference fluid. Accidentlly the reading was 29.6 psu stayed within about 5 minutes. I took it back to the sump, and the reading changed 30.1 psu from 29.6 psu.
I presume that this change might be due to the cleaning in rodi water. After 8 hours, and it still shows same 30.1 psu now. I will keep watching the change.
I understand that all the instruments can not have accurate same readings, especially in case of conductivity instruments even after calibration, but what I expect is stable readings with some tolerance. All the other probes are very stable and accurate. I periodically recalbrate them, and I measure the refernce fluid before each calibration, but drifts are negligible.

FYI, my tank is as follows.
600L sps dominant reef tank, 34.0+/- 0.1 psu, Alk 136 +/- 2 ppm, Ca 420 +/- 5 ppm, Mg 1350 +/- 5 ppm, medium/low bioload, low nutrient system running with a calcium reactor w/ an alk dosing pump and a macro algae reactor. I don't see any sign of non-bio calcification. Salinity is measured by Hanna HI98319 and Veegee STX-3 calibrated with a calibtation fluid. also the calibration fluid is verified with two more calibration fluids from different manufactures. All others are measured with Salifert test kits.
In the meantime, does ozone affect the probe sensitivity? I have been running an ozonizer with my protein skimmer since about two to three weeks ago.

Is there any other measure to find the reason or to get it back to stable measurements?

Son

Please recalibrate the probe.
After calibration, let it sit in the fluid. It should read 50mS or 32.7 PSU. Once verified, increase the signal filtering by (1) or (2), then observe over the next few days.

OK Vinny.
I will be doing a calibration as per your instruction.
I will be back after watching the trend for 5 to 7 days.

Son

Hello Vinny,
Long time passed, but I resolved this problem by adding an industrial purpose isolation transformer in the power supply to the controller of my Royal Exclusiv RD3 DC skimmer pump motor.

The followings are feedback for your future reference.
I calibrated the salinity probe as per your instructions and measured the tank water in a separate container on Dec. 24, 2019. The readings were 32.1 psu after a calibration with 50mS solution and up and down between 32.1~31.7 psu for 2 days. I assume the reading were normal and stable, a little fluctuation was due to ambient temperature changes.
I took the probe back into my sump. The readings were 32.5psu and continuously drifted down to 0.2 psu in about two weeks (on Jan. 9, 2020).
I judged that something in the water column disturbed the salinity measurement.

Some time ago, couple of local reefers complained that their incoming ELB (Earth Leakage Circuit Breaker) tipped whenever they ran their R.E RD3 DC return pump or skimmer pump. The C.S of the manufacturer insisted their grounding was poor. (I understand this is not true.)
As we have 220V 60Hz electrical supply system, so many R.E users add a 220/110V step-down transformer with US version RD3 pump. I recommended to replace their single winding step-down transformer with a double winding isolation transformer. They all resolved their ELB trip problem.
I assume the RD3 motor block had some leakage current due to degraded insulation thru their operation for an extended period. If the SMPS for the RD3 had such perfect insulation capability, the leakage on DC side could not flow to the grounding point (normally neutral of star connection) of the local distribution transformer to trip the ELB.

I assumed, in my case, the leakage current was not enough to trip the ELB, but enough to disturb the measuring of the conductivity probe. Also in my understanding, if the galvanic isolation circuit of the conductivity probe in Profilux 4 and/or the isolation (insulation) capability of the SMPS for the controller are sufficient to block the leakage current, the leakage current loop will not be established. I do not complain about the circuits of profilux 4. It’s a meter of reliability of domestic, industrial, automobile or aerospace grades of circuitry. But the long-term insulation reliability of RD3 is something suspicious.

I think also this kind of problem happens because every country has different grounding system or different ground current detection system.
I don’t mind the actual measurement of the probe after a calibration, I can give some offset value on the calibration referring to a reference from GMH3451 with lab grade calibration solutions. Important point is measuring stability and the current readings of my probe is fixed on 32.2 psu for three days.
I appreciate the stability of the measurement and I'm happy.
Have nice day~!

Son

Hello Son,

Thank you very much for the update.
That is an interesting observation which will hopefully help others with the same pump who are also experiencing the same issues.