This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMP91000: Inquiry for alternative solution to LMP91000

Part Number: LMP91000

Tool/software:

Hi, TI expert

Customer's products are as follows:  Industrial Portable Multi gas detector

The customer wanted to apply LMP91000, but is inquiring about alternative solution.

As introduced in the datasheet of LMP91000, the programmable features of LMP91000, such as adjustable bias voltage, adjustable gain resistance, and adjustable load resistance, were intended to be applied as an advantage to support multiple types of electrochemical sensors such as 3-electrode toxic gas sensors and 2-electrode galvanic cell sensors with a single design.

However, due to the issue of not being able to use the programmable component during the intrinsically safe explosion-proof certification process, it is expected to be difficult to apply the LMP91000.

It became necessary to find a solution consisting of a non-programmable component that can replace this advantageous function.

Q1) Is there a solution that can be designed to apply multiple types of electrochemical sensors in a single design without setting registers (non-programmable components)?

Please check. Thank you.

  • Grady,

    If you want to avoid the programming of LMP91000, the customer can recreate the analog components of the circuit based on the block diagram and the bias specifications of their electrochemical cell.

    At a bare minimum, the customer will need to bias the electrochemical properly (usually 10s or 100s of mV, positive or negative), and implement a transimpedance amplifier on the WE output to convert the output current to a voltage. Since the LMP91000 offers all of this integrated into the chip, we don't have a specific recommendation on what TIA to choose, and the other details such as the biasing and TIA gain will depend on the customer's system.

    Regards

    -Alex Thompson

  • Hi, Alex Thompson

    I have one more inquiry from customer regarding driving LMP91000 Lead Free Type O2 sensor.

    The customer is currently reviewing operation of the Lead Free Type O2 sensor by applying the LMP91000 setting values ​​as shown below.

    ** TIACN : 0x13, REFCN : 0xeb, MODECN : 0x3
    - VREF 3.0V
    - VREF DIVIDER 11 (Bypass) -> 3.0V
    - VARIABLE BIAS 20% -> 0.6V
    - RLOAD -> 100Ω
    - RTIA -> 14KΩ

    Can you please confirm that the settings are correct to apply Electrical Bias Voltage -600mV?

    Please check. Thank you.

  • Grady,

    These look like good settings to me. Please let me know if you have any issues getting the desired bias voltage to the control electrode.

    Regards

    -Alex Thompson

  • Hi, Alex Thompson

    The customer is driving the Lead Free O2 sensor by applying the LMP91000 register settings as below, but the sensor response is not working.

    - VREF 3.0V
    - VREF DIVIDER 67% → 2.01V
    - VARIABLE BIAS 24% → 0.4824V
    - RLOAD → 100Ω
    - RTIA → 14KΩ
    ** TIACN 0x13, REFCN 0xcd, MODECN 0x3

    You can also find the E2E link contents of the current situation similar to the current situation below.

    https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/1456630/lmp91000-i-am-having-issue-while-configuring-o2-3-lead-sensor-with-lmp91000?tisearch=e2e-sitesearch&keymatch=lmp91000%20o2%20sensoe#

    Can you please check what the problem is?

    Please check. Thank you.

  • Grady,

    Thanks for the detailed response. Is the customer using the LOCK Protection Register (Address 0x01) to make the registers R/W, and not just read mode? Please make sure they are setting that register value to 0x01, to unlock the LMP91000.

    I think the bias voltage set to the CE pin of the O2 sensor is not correct. I would try setting the internal zero divider to be bypassed (so no division), and then use a 20% bias setting. With an external VREF of 3V, 20% of that will be 600mV (still keeping it negative). So have the customer write 0xEB to the REFCN (address 0x11), and see if that improves their results.

    Regards

    -Alex Thompson