• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 31 subscribers
  • Views 561 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

LMP91000: Current-to-Voltage Amplifier (15na) Amperometric, 3-electrode sensors

LMP91000: LMP9100 help for setup and strange behavior

Rey Cédric
Prodigy 30 points
Part Number: LMP91000

Hello,


I'm coming back to you regarding the LMP91000 chip. I can communicate with it but I can't adjust it correctly. I've read and reread the LMP91000 datasheet several times but I don't know if it's my level of electronics or my level of English that's lacking.

I have a VOC sensor (ES1-AG2000). Its data sheet mentions that it's an Amperometric, 3-electrode sensor. I've connected it correctly. According to the data sheet, PPM values are converted on average in 25nA increments. Zero Current is +/- 100nA. The recommended load resistor is 100 Ohm, which is exactly what the LMP91000 offers.

On the other hand, it says Bias Voltage = 0mV. And I may have a problem understanding the definition of "Bias".

So in theory, when I connect the sensor to the LMP91000 and there's no pollution in the air, I should have a value of 100nA. With a 14kOhm TIA_GAIN resistor, I should read a value at VOUT of around 1.4mV (V = I * R). Knowing that my load cell has a capacitance of 2000ppm, the maximum would be 50100nA or 701.4 mV with the same load resistor.

My configuration should therefore be :
TIACN
TIA_GAIN = 100 14kΩ
RLOAD = 11 100Ω

REFCN
REF_SOURCE = 0 Internal (I have a VREF of 2.5V too, but I don't see any difference by changing this parameter) Otherwise VDD = 3.3V
INT_Z = 00 20% / I would have said 11 for bypass, but obviously it's only for sensors with 2 pins.
BIAS_SIGN = 1 Positive (VWE -VRE)>0V (I imagine)
BIAS = 0000 0% (default) because the sensor indicates a BIAS of 0mV

MODECN
FET_SHORT = 1 Enable (I would have said it should be enabled, as it's stated in 7.4 "The zero bias between the WE and RE electrodes is kept by
enabling the internal FET feature".
OP_MODE = 000 Deep Sleep (default) or 011 3-lead amperometric cell

Here again, Deep Sleep could be used, but it doesn't amplify the current... or 3-lead amperometric in standard configuration, which also supports the amplifier.

I've tried lots of variants, but no real success! With the following parameters :
LMP91000_TIA_GAIN_350K | LMP91000_RLOAD_100OHM,
            LMP91000_REF_SOURCE_INT | LMP91000_INT_Z_20PCT
                  | LMP91000_BIAS_SIGN_POS | LMP91000_BIAS_0PCT,
            LMP91000_FET_SHORT_DISABLED | LMP91000_OP_MODE_AMPEROMETRIC

I'm reading 1.88mV, which is good, but only because no reaction occurs when I expose the sensor to VOCs. However, if I change the TIA_GAIN, the value changes. At 7K I read a value of 13.13mV, but again no change!

If I switch to LMP91000_OP_MODE_DEEP_SLEEP (0x00), I read around 580mV. On the other hand, changing the TIA_GAIN doesn't seem to change any parameters. In this setup, I can still see a reaction from the sensor, I've seen it go up to 800mV with the right saturation. But what's the relationship between mV and PPM? The reading seems to oscillate up and down quite a bit.

LMP91000_TIA_GAIN_14K | LMP91000_RLOAD_100OHM,
            LMP91000_REF_SOURCE_EXT | LMP91000_INT_Z_50PCT
                  | LMP91000_BIAS_SIGN_POS | LMP91000_BIAS_20PCT,
            LMP91000_FET_SHORT_ENABLED | LMP91000_OP_MODE_AMPEROMETRIC

In this configuration I'm at 3200 mV! However, I don't feel that the sensor is reacting.

If someone can help me to understand because I'm lost with all this.

Thanks in advance,
Cédric

  • 0
    TI__Genius 9485 points

    Hi Cedric,

    Thanks for reaching out again. The voltage that you are reading that won't change, is that on the Vout pin? I would also agree that based on the datasheet for the sensor, you shouldn't need a bias voltage. But you might want to try one to see if that helps.

    You can also try to download the PSPICE model on the LMP91000 product page to try and simulate different configurations.

    I will continue to research this as we go, please let me know if the voltage you are measuring is on VOUT.

    Thanks

    -Alex Thompson

  • 0 in reply to Alex Thompson
    Prodigy 30 points

    Hello Alex,

    Thanks for your reply. I'm ashamed of myself... Last night I found the problem, which was really a problem of understanding and interpretation!

    The bias voltage is one thing, it must remain at 0 for the sensor. However, I shouldn't have mixed it up with INT_Z, which corresponds to the zero voltage of the LMP91000 and is mandatory!

    Clearly, it's perfectly logical to obtain a value of 500mV with 20% INT_Z, since my external reference current is 2500mV. And this value cannot be bypassed! Of course, the LMP91000 will never start at 0, and that's what you have to take into account.

    To go back to my example, and in the hope that it may help others who didn't understand the subtlety like I did:

    If you've set INT_Z to 20% to 50% or 67%, it's a mandatory value that will define the minimum VOUT output. So by selecting 20% of 2.5V, we'll read 500mV in any case. You'll never see a lower value unless you change your VREF value, but that's not the point. This has no effect on the sensor! Then, below this base value, considered as the zero of the LMP91000, comes the zero of the sensor. In my case, after adjusting the 12kOhms resistor, we can expect an additional 12mV equivalent of the 100nA.

    So the basic formula is :

    PPM = ((Voltage read in mV on VOUT - INT-Z in mV - ZeroSensor in mV) / PPMRatio)

    INT_Z = 500 //example in the case of 20% of 2.5V on V_REF
    ZeroSensor = 12 // 100nA converted with a 12kOhms resistor
    PPMRatio = 3 // 25nA on average converted with a 12kOhms resistor (25 * 12) / 100

    Then, in a 2nd step, we can plan to vary PPMRatio as a function of temperature, according to the sensor datasheet!

    Now I've got everything I need to make it work, and I hope it can help others with the same problem in the future.

    Best Regards,
    Cédric

  • 0 in reply to Rey Cédric
    TI__Genius 9485 points

    Cedric,

    Please don't feel ashamed! The LMP91000 is not a trivial device, and I'm really happy to hear that you were able to make this work. I'll go ahead and close this thread.

    Thank you

    -Alex Thompson