CN0359 conductivity strange problem of reading values

Hello...

We are using CN0359 for measuring conductivity. We have 2 calibration liquid . ( 1.4 mS solution and 12.9 mS solution )

We use 4 graphite electrode sensor from another company. The sensor works well with its control unit.

When we plug the sensor to CN0359 we calibrate the K value witch is 0.475

Our Ecx Voltage : 2 V

Freq : 100 Hz

Hold time : 10

Set time : 10

The strange problem is when we put the sensor in to solution and power the CN0359 the reading value is successful.

For example when we use 12.9 mS solution the reading is also around 12.9 with good accuracy.

Our problem is when we put out the sensor from the solution The EC is getting close to nS which is good also.

But when we put back it in to solution we are not getting 12.9 mS again from the reading.

When we put the sensor inside the solution and power off and on again the CN0359 the Ec come to 12.9 mS againg with no problem.

We observed the Gain Values from the ADC inputs off the EC inputs.

When we put back the sensor in to solution Vgain become to normal value but I gain is not turning to same value.

I gain stick in to area that there is no solution.

When we want to make good measure we need to put to sensor in to solution before we power on the CN0359.

Do you have a solution for us to fix this problem. 

Regards...

  • 0
    •  Analog Employees 
    on Oct 16, 2019 1:15 PM over 1 year ago

    Hi Raseyns,

    I will look into this and replicate your setup on my end. Can you also give me the model of the conductivity sensor?

    On your end, do you have a variable resistor around? A simple test you can do is to use a variable resistor as the conductivity cell and wire it in a 4-wire config. With the CN0359 powered on, you can set the variable resistor to a value of 50 or 100 ohms, note the conductivity measurement, set it to a very high value, then check the measurement to see if it reads a very low conductivity. If it does, set it back to 50 or 100 ohms to see if it is still able to get the same measurement as the first one.

    Regards,

    Nikko

  • Hello Nikko Thanks for your reply.

    We have 2 CN0359 kit. It shows that one of the kit makes this problem but the other one works fine.

    So we think that maybe we damage the the hardware lilttle bit.

    Anyway thanks for reply.

    But I have a new question for developers.

    In our 2 CN0359 we observed that there is a warm up in one IC.

    The warming IC is ADA4627-1ARZ

    Its the closes IC to conductivy sensor connector ( pin j5)

    Why this IC is warming high ?

    Can we stop this warming ?

    We want to use CN0359 design as a reference to our design.

    So instead of ADA4627 can we use another IC to stop warming up?

    Regards....

  • 0
    •  Analog Employees 
    on Nov 18, 2019 12:26 AM over 1 year ago in reply to raseyns

    Hi Raseyns,

    Sorry for the delay. The warming up of the ADA4627 is due to its dissipation which may cause it to be 20 to 30 degrees Celsius above ambient. The supply current for the ADA4627 is typically at 7mA. This has negligible effect on the measurement since the offset drift is at 2uV / degrees Celsius which ensures the drift voltage is minimum for the entire temperature range.

    If you really need to remove the warming effect of the IC, I suggest to use low power low input bias current op amps such as the LTC2064.

    Regards,

    Nikko

  • Hello Nikko.

    Thanks for reply again. 30 Degrees above the ambient may accur a problem. Not sure abouth it. So its better to take a B plan.

    The LTC2064 IC that you suggest is working with Max + / - 2.5V which is not enough for our circuit. Because we will use +/- 15V for power supply to that IC.

    Can we fix it in another way ?

    Regards...

  • 0
    •  Analog Employees 
    on Nov 20, 2019 3:06 AM over 1 year ago in reply to raseyns

    Hi Raseyns,

    You can check the LTC2058 and LT1462. The LTC2058 has about 1mA supply current, 30pA input bias current and 0.5pA/sqrt(Hz) noise current spectral density. The LT1462 has 22uA supply current, +/- 1pA input bias current, and 0.5fA/sqrt(Hz) noise current spectral density. Both are capable of +/- 15V supply voltage.

    Regards,

    Nikko