LMP91200: Using LMP91200 with other sensors in the same Liquid possible interference

Hello Armrit,

Could you give a little more detail on how your PH readings are becoming inaccurate and include your schematic? That will help me to rule out other potential issues.

Thanks,
Audrey

Hi Audrey,

Just to elaborate, the value would drift for PH. For example if I placed the probe in a water of PH 7 after sometime it would drift to a higher PH value (ADC ticks would decrease). This would happen fairly shortly within half an hour.  Then when I moved the probe to another solution of lower PH that would also read higher PH values almost as if the voltage was being biased by something in the water or the circuit itself. Once i put the probe back into calibration solution it seems to reset its self back to what PH 7 and PH 4 used to be before the drifting.  The problem is not with the Probe we buy high quality dual junction probes from BlueLabs and have used them in other PH circuits with fairly good success.

Why I was thinking about an isolator is that we have a EC circuit which generates a current through the water VIA the EC probe. Whenever this is running the PH seems to go way off. Like reading 12 when it should be 4.  Clearly a ground loop is generated from this which is effect the LMPs ability to take readings.

I have attached my schematic below and have followed the reference design for the most part. The dashed line repersents the seperation between digital and analog GND planes.

Hello Amrit,

Thanks for the additional information!

After consulting one of our TI Reference Designs, TIDUA47B, I would suggest you check a couple of things: 1) the temperature effects and 2) the layout guidelines. With regards to temperature, the LMP91200 has a pin dedicated to connecting an RTD for temperature compensation because the pH electrode varies sensitivity across temperature (see Fig. 5 in the reference design). I noticed you do not have an RTD in your schematic; if you have temperature variations in your solution you may expect to see some variation in the output like the Reference describes. Secondly, TI suggests some guidelines to reduce the parasitic leakage current using guard conductors surrounding the input line (see Section 5 in the reference design). This helps to isolate the LMP91200's input from large voltage gradients across the PCB.

We do not have any references addressing the use of a digital isolator in conjunction with the LM91200, so I cannot confidently say whether or not that would solve your problem. I would suggest you contact the sensor manufacturer to get more information on the implementation of the pH sensor in your application. Our references and WEBENCH tools are focused on the AFE implementation with the sensor.

Regards,

Audrey

Hello Amrit,

There has been a change between the LMP91200 device shown in the reference design and the current version. Some features, including the RTD connection, are no longer offered by TI. I apologize for pointing you to this reference before noticing this discrepancy.

However, we do offer an RTD replacement device, the TMP117, which is a highly accurate digital temperature sensor. If you find the temperature is indeed affecting your sensor drift, then you can use the TMP117 to interface directly to your microcontroller to compensate.

Regards,

Audrey

Hi Audrey, I do use the guards rings as suggested in the reference design. I don't believe temperature is an issue as I calibrate and test the water at room temperature. Do have any notes on susceptibility to noise/ground loop or any recommended PH probes which would work with this device. Is there anyway that my ADC could effect the readings.

Thanks