I am using the CN0349 to measure conductivity in seawater.
However, the problem is that after calibrating, performing a sweep, calculating conductivity, the conductivity values tend to drift over time. It drifts over a matter of seconds.
I have made my own stain less steel probes to sit in a solution.
For example, my received conductivity values will start at 4.9 and every second decrease by 0.003. There seems to be no end in sight either, as it doesn't seem to settle.
I have 2 settling times
start frequency of 100khz
and I just repeat the frequency measurement five times
Excitation Voltage 2.0Vp-p, Internal PGA=1
The cell constant of the probe was calculated to be around 40
The measurement uses rfb = 100ohms
any help is appreciated.
This is not a problem when measuring the impedance of a resistor. Only in a saltwater solution
This is a tricky situation, because you are saying that the board works fine and is consistently producing the correct results when using a precision resistor. So that leads me to believe there is either something going on with the probe you made or the seawater solution you are trying to measure (or even the combination of both).
The first thing that I would try is to use a commercial conductivity probe and see if that gives you reliable results in the sea water solution? This will tell us a if your probe is the thing causing the errors. I'm thinking that there is likely going on with your probe, although it is strange that it never settles out.
What software are you using? Are you using the Labview GUI that ADI provides? Or is your own software that you have written.
I am using C code that I wrote myself to calibrate, save the gain factors on an MCU EEPROM (arduino), and take measurements for temp and conductivity. I am taking measurements like every second. Over maybe a minute, the measurements will drop consistently or increase consistently.
I thought the issue could have been my specific CN0349 so I purchased another one and to my dissatisfaction saw the same trend.
Using the ADI software was too slow to see much of a drop or increase. It was hard to tell if the conductivity was ramping up or down because the values were coming in at a lot slower rate. In general it seemed to neither ramp or slope down, but stay evenly noisy in both directions. This does not mean that it doesn't exist only that I cannot see it.
I also snubbed my sampling frequency down to see if that was it: it was not the issue.
Taking the probe out of the water "zeroing" the measurement and placing it back sort of restarts the process of getting a value and then drifting in a direction. I am using a pretty big bucket of seawater so the results might make sense. However I have a sea bird conductivity CTD in the bucket as well. And it doesn't indicate any drift in the solution. If I make a change to the solution, like add more salt or dilute with water, both the seabird and the CN0349 can display the drop or increase.
I have tried using different start frequencies, increments, frequency resolutions, and settling times,as well, to no avail.
My probe is two stainless steel pins looking into the water, epoxied into an enclosure. I calculated a cell constant of 40 I think. I also I don't have a commercial probe to test with, so to verify its not the probe, I am limited to various fixed geometry plugs that I have laying around. I used a BNC cable and a two pinned phone jack as psuedo probes.
However the effect is the same with every probe, drifts up or down. I ordered a commercial probe to test this on, but it hasn't shipped yet.
The temperature probe seems to be ok and measuring about .8 degrees off what the other instrument says.
The only thing I can think of is when I reupload my code to the arduino, without changing the orientation or solution. The output values change by a little. Is there a reset I can do on the CN-0349 to emulate this?
Here are some graphs of what I have done so far. The first photo shows 800 data points taken every second or so or steady state solution. On the commercial instrument though it is level and near the end increasing.
This photo shows data logging of both increase and decrease. The first two graphs are time periods of over hours so thats why it looks smoothed out. However you can see the negative gradient. The third graph shows the positive gradient. TThe bottom middle graph shows when I increase the salt content in the solution. It goes up but slowly decreases.
The problem is more that if this drift is small, it could turn out to be really big over hours when I want long term monitoring
I still think we need to step back and look at the bigger picture here.
Could you connect a known impedance(precision resistor) to the input of the CN0349 and take the same data you were taking with the probe/seawater? Can you see if this drifts as well?
If this test doesn't drift then I have to think something is wrong with the probe/seawater setup or perhaps even in the software. Could there be a run away issue where you are adding a constant to the measurement within the software?