• TI thinks resolved

Dual pH sensor LMP91200 to ADS1248, isolation is needed?

Hello everybody!

We are working on the development of a dual pH and dual Pt100 sensor interface for a fish farm tank control system.

The tank is expected to have about 75m3 of water, a number of pumps, and feeding systems that are expected to inject noise.

Since the distance between the two pH sensors is expected to be about 6m is it needed an isolation between the signal conditioning cirquits of the two sensors (like the attached cirquit)?

In the attached files we are two isolation approaches and the signal conditioning cirtquit for one sensor for the approach #1.

Thank you in advance,Isolation #1 & #2.pdf

Tank_Interface _ IN_pH.pdfNikos


  • Nicolaos:

    I'm NOT a TI employee or related in any way to the company but if you want some free advice ... :)

    I'd suggest you go for design option #2 for the simple fact that its easier to isolate digital signals with no issues than analog signals. There are plenty of good digital isolators that exists where you're digital signal will be transmitted unmolested by noise and RF interference.

    Analog will not fair well in the presence of noise especially on long cable runs, even when you have a differential input configuration into your ADC to try to cancel out common mode noise. Pumps are extremely noisy and are well known RF noise polluters.

    The AMC1200 adds more noise to the signal via extra bias, voltage offsets and distortion.

    Its much easier to clean up a digital signal than an analog one.

    Just my 2 cents...

    Good luck!

    - David Musoke

  • In reply to DMM:

    Dear David,

    Thank you very much for your advice and your support...

    We have designed our prototypes isolating analog parts in-between them and in-between our SBC (pretty much close to your advice).

    We are waiting for the board production to assembly them and start the testing...

    Best regards,

    Nikos Piliouras