LMP91200 - pH measurement not match mV equivalent

Hi Miguel,

is the liquid under test at ground potential ? or is the container of the liquid under test in metal and grounded ?

The LMP91200 forces the liquid under test at VOCM potential one a PH probe is connected to the LMP91200. If the liquid is already grounded there will be an error in the reading.

regards,

Domenico

Hi Domenico,

Thank you for your answer.

The liquid under test is in a glass counter. It is not at ground potential.  Here is a copy of our pH input PCB layout:

Please let us know if it is correct. We followed the specs on the datasheet of the LMP91200 to removed the solder mask around the input lines of the LMP91200 and added the guard ring on both the top and bottom layers of the board.

Regards,

Miguel Lopez

Hi Miguel,

the layout suggested in the DS is different, the connector showed in the Datasheet (below) is a triax. The main central connector of the Triax is INP, the second central connector is the guard, the shield connector is the VCM.

Your layout even if different (based on 2 pin connector) is good.

I have some doubts on the VCMHI. Where have you connected that pin ? it is a very sensitive pin (high impedance) so you should leave it floating, any load connected to that pin generates a drop on VCM.

regards,

Domenico

Hi Domenico,

We connected the VCMHI pin to a test point. The voltage in the VCM pin is 1.25V and Vref is 2.5V.

We did more tests to try to understand our problem. We added a voltage buffer between BNC connector input and the INP input of the LMP91200 and that solves our problem. The error is now +/-0.02pH after calibrating our devices using a mV source and measuring pH buffer solutions for 4,7 and 10pH using an electrode. Without adding the voltage buffer, the error is +/-1pH after calibrating our devices using a mV source and measuring pH buffer solutions for 4,7 and 10pH using an electrode.

The LMP91200 should work without a voltage buffer connected at the INP input. Could you please tell us if adding the voltage buffer is necessary or we should look for another solution?

Regards,

Miguel Lopez

Hi Miguel,

the LMP91200 does not need any buffer at INP pin. The core of the LMP91200 is a buffer with very low bias current (tens of fA).

The device has been tested with several pH probes of different manufacturers.

Which PH probe are you using ?

regards,

Domenico

Hi Domenico,

We are currently using the PHE-1311from Omega Engineering : http://www.omega.com/pptst/PHE1300_1400_2300.html

What else could be causing the mismatch we are experiencing with the pH measurement?

Is there any other board layout we should take on consideration that is not mention in the datasheet? In the development board for the LMP91200, we see that Guard 1 and guard 2 are, apparently, not connected together on both layers of the board but it is recommended on the datasheet. Could you please clarify for us?

Regards,

Miguel Lopez

Hi Domenico,

Any updates? Did you look to the specs of the PHE-1311?

Regards,

Miguel Lopez

Hi Miguel,

I've never tried a filled gel ph probe, which is the difference between a gel filled pH probe and a ph probe with not gel reference liquid ?

Is there any chance that a gel filled probe has huge output resistance ?

regards,

domenico