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.

LMP91200: Low-cost Opamp for PH sensor

Part Number: LMP91200
Other Parts Discussed in Thread: OPA391, LMP7721, LM4140, OPA928, TL431, TL432LI, TL431LI, LPV821, TL432, LM4040, REF35

Hi Ti-engineers.

We have a product used PH sensor, we designed and used with LMP91200.

But now we need a low-cost solution.

Could you please help to check and give me an advice for Opamp and Vref ?
ex: analog output range (0.5-1.8V)

Thanks for yours support.

  • Hi Alone,

    But now we need a low-cost solution.

    Could you please help to check and give me an advice for Opamp and Vref ?

    Please provide me with the additional information.

    Are you good with the existing LMP91200 pH sensing performance? 

    Are you continuing to use LMP91200, but you want to low cost in some of op amp and Vref design?

    What is the target cost for the similar pH sensing performance?

    Are you willing to go with total discrete pH sensing solution?

    I need the above information so that I can better assist you. For instance, OPA391 may work for PH sensing application, but I'd to know where the cost saving is going to be (without degrading the sensing performance). 

    https://www.ti.com/product/OPA391?utm_source=google&utm_medium=cpc&utm_campaign=asc-amps-null-prodfolderdynamic-cpc-pf-google-wwe_int&utm_content=prodfolddynamic&ds_k=DYNAMIC+SEARCH+ADS&DCM=yes&gclid=Cj0KCQjw5f2lBhCkARIsAHeTvlgFLSvDK5lFYmbbvWdeY4Fb3jXCHr9fZtRIqVPaAS4LBZgPjJBnvfkaAjPrEALw_wcB&gclsrc=aw.ds#order-quality

    Best,

    Raymond

  • Dear Alone173 - 

    I would also add here to Raymond's input this reference design using the LMP7721 & LM4140 (precision reference you were looking for, too): https://www.ti.com/lit/an/snoa529a/snoa529a.pdf 

    then compare to this reference design, using the LMP91200: https://www.ti.com/tool/TIDA-00561 for your choosing the right solution for you overall. (cost/complexity, accuracy, ease of use, etc.)

  • Hi Raymond.
    Thanks for your feedback.

    Are you good with the existing LMP91200 pH sensing performance? 

    => not good, still not stable when reading adc, i think main issue is vref and dont have connection for VREF (config at 1.5V, AZ431LBNTR-G1) of LMP91200 to MCU.

    Are you continuing to use LMP91200, but you want to low cost in some of op amp and Vref design?

    => yes, you're right.

    What is the target cost for the similar pH sensing performance?

    => under 2.5-3$ for both vref and op-amp ic ( 1000pcs).

    Are you willing to go with total discrete pH sensing solution?

    => not sure, but i used external PH probe.

    OPA391

    => its seem cost too high with GP op-amp.
    Could you show the highlight of this part adaptive for my product?
    Thanks.

  • Hi Josh Wyatt
    Thanks for your recommendation.

    I saw this application note and thinking about simple solution to merge both U1 and U2 ( op-amp) .
    Also, i'm finding all parts with low cost as possible.

    Thanks.

  • Hi Alone,

    => not sure, but i used external PH probe.

    Most mid to high end pH meter is equipped with external pH probe. Does your pH probe comes with Ag/AgCl reference inside the probe? If the pH probe has the chemical reference, then you can use it as the reference (rather than electronic reference). 

    0083.PH_probe_design_national_AN-1852.pdf

    => under 2.5-3$ for both vref and op-amp ic ( 1000pcs).

    In our pH application note, we suggested to use LM4140A-1.0 voltage reference (1.024Vref), which it has 0.1% accuracy (grade A) with 3ppm/C drift. Yes, this cost more than others. The circuit is shown below. 

    When the Vref is divided down, the buffer U1 is measured at 1.024/2 = 512mV. As long as the 512mV or the reference is stable, then Vout will be stable as well. 

    In the Nernst equation, the key to accurate pH measurement is T (degree K = 273.15 + C), op amp U2 & U1 and ultra stable and low noise V+ supply voltage. 

    Here is what I would do in order to lower the cost and improve the pH measurement performance. 

    1. Get a precision RTD temperature sensor for the pH measurement solution. Sometimes, the temperature sensor comes with the pH probe, which it will be good, if accuracy and repeatability is good. 0.01C to 0.1C temperature accuracy for the application. These days, RTD sensor is widely available and should not cost much. TI has many RTD and accurate temperature sensing application note. I can send these to you, if you are interested.  

    2. If you use Ag/ACl chemical reference, then you do not need electric potential reference. The standard electrode refers to Ag/AgCl electrode in saturated KCl solution, which its electrical potential is measured at approx. 198mV. 

    3. If the pH electrode does not Ag/AgCl chemical reference, then you need to have a fairly good bandgap reference + U1, which zero drift precision op amp buffer + electrical voltage reference.  I may be able to find you a suitable electrical voltage references. 

    4. Is your system battery powered or use regulated LDO? Please make sure that V+ is ultra stable and low noise. Please let me know.  

    5. You may continue to use LMP91200 or use discrete LMP7721 (OPA928 is our latest and greatest low Ib op amp, but it costs more). If you happy with LMP91200 IC, then continue to use it. LMP7721 and other ultra low Ib op amps may work as well. Since most of pH probe is high impedance device, therefore ultra low Ib op amp is required. If you tell me the impedance ranges of your pH electrode, I can find an alternative op amp, if you want to go with discrete solution. 

    6. Last thing is the cable length and shielding. If you purchase these from a manufacture, then it should be adequate. Please make sure that you have a good electrode contact with the precision op amp. 

    Once we have these issues answered, we can simulate the entire pH electrode and check out the electronic noises and its performance. 

    If you have other questions, please let me know. 

    Best,

    Raymond

  • Hi Raymond
    Thanks for your details.
    Finally, could you show the part number for both Vref and Opamp that meet my target cost ?
    Thanks.

  • Hi Alone,

    could you show the part number for both Vref and Opamp that meet my target cost ?

    Here is the low cost option. I would select ±0.5% B grade or better in TL431 reference. The reasons for the selection is that the Vref for pH measurement is a just an electrode chemical reference, it can be Ag/AgCl or electronic one. This reference has to be placed very close to the pH measurement electrode. As long as it does not change between calibration and measurements, it should be good. Ultimately, the Vref is dropped out from the pH calculation. 

    You should spend more investment in temperature measurement. With 0.1C to 0.01C of the temperature accuracy of the measurement pH solution, the accuracy will also improves. 

    pH Vref with LPV821 07312023.TSC

    Here is the search list of low cost voltage references: For instance, TL432LI may be better than TL431, which TL431 is for higher current application. T432LI or TL431LI, 0.5%, B grade package would be better option than TL431 for pH measurement application.   

    https://www.ti.com/power-management/voltage-reference/products.html?login-check=true#2352max=0.025%3B0.5&sort=1130;asc&

    If you have other questions, please let us know. 

    Best,

    Raymond

  • Here is the low cost option. I would select ±0.5% B grade or better in TL431 reference. The reasons for the selection is that the Vref for pH measurement is a just an electrode chemical reference, it can be Ag/AgCl or electronic one. This reference has to be placed very close to the pH measurement electrode. As long as it does not change between calibration and measurements, it should be good. Ultimately, the Vref is dropped out from the pH calculation

    Yes, i also used the part number same that TL431 (AZ431LBNTR-G1) in my design.

    But the result have not stable when reading ADC from LMP91200.
    On the current PCB, we place a vref IC as far as with PH probe connector, so maybe need adjust in the next version.

    You should spend more investment in temperature measurement. With 0.1C to 0.01C of the temperature accuracy of the measurement pH solution, the accuracy will also improves. 

    Yes, we have a RTD sensor on board to monitor the temperature.
    We will note to recheck it.

    Its seem your final solution is using TL431(VREF) and LPV821 ( opamp).
    It will be better if you could help to check on the prototype at TI's Lab and give me a report.

    Many thanks for your help.

  • Hi Alone,

    Yes, i also used the part number same that TL431 (AZ431LBNTR-G1) in my design.

    AZ431LBNTR-G1 is NOT TI part number, and you should ask for Diode Corp for the support. In addition, I request to get ±0.5% B grade, you buy ±1% part. What do you want me to do?

    Best,

    Raymond

  • hi Raymond
    The part number AZ431z was used on old board.
    I want to check before replace the new part which you recommended.

    Please note that I want to find a part number have more stocks to order.
    Thanks

  • Hi Alone173,

    In the previous reply, I attached the following link in TI E-store. If you login your user information, you should be able to see the inventory availability on various parts, pricing, performance etc.

    https://www.ti.com/power-management/voltage-reference/products.html?login-check=true#2352max=0.025%3B0.5&sort=1130;asc&

    You should try out the following voltage reference, again 0.5% or better. 

    https://www.ti.com/lit/ds/symlink/lm4040-n-q1.pdf?ts=1691164501181&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FLM4040-N-Q1

    https://www.ti.com/lit/ds/symlink/tl432li.pdf?ts=1691163839281&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FTL432LI

    TL431LI / TL432LI Programmable Shunt Regulator with Optimized Reference Current datasheet (Rev. A)

    https://www.ti.com/lit/ds/symlink/ref34-q1.pdf?ts=1691126150010

    I do not know what the budget for your references, if you are able to get Low Temperature Coefficient ~100 ppm/°C, which it would be better choice. The electronic reference in pH measurement application does not require the absolute accuracy, but it requires low temperature drift and consistent voltage reference. You need to understand that you get what you pay for. Your electronic reference selection should be based on the performance/price ratio for the application. 

    Best,

    Raymond 

  • Hi Raymond
    Thanks for your detailing.
    Really lowest cost is most requirement in this version update, but efficient and stable is also important.

    As your recommend, I will try to testing with LM4040, TL432 and LPV821 ( maybe better if you can recommend a lower cost part)
    Also, can i get a sample for above parts for testing?
    Thanks

  • I also try to tested with your circuit in Tina.
    its seem you set reference voltage to 250mV, is it too small to make offset voltage ? (Typical Ph range is +/-0.4V).
    Thanks

  • Hi Alone173,

    can i get a sample for above parts for testing?

    The application team does not handle the sample request. If you are interested the samples, please follow the following link or you may contact our FAE in your area. 

    https://www.ti.com/ordering-resources/buying-tools/sample-request.html

    its seem you set reference voltage to 250mV, is it too small to make offset voltage ?

    If you are using the chemical reference, which is a standard pH measurement, the typical reliable electrode potential is approx. +198mV, depending on the Ag/ACl reference physical configuration. 

    The LMP7721 application note is using +512mV as the electronic reference, and you may use +400mV as well. As I implied, the actual electronic reference is not critical, as long as it is stable and unchanged during the pH calibration period. Sometimes, if Vcm of an op amp has to be operated 200mV or above the GND potential (say in a single supply), +400mV or similar reference potential range will make more sense, but this is to optimize the op amp's bias voltage for the electronic measurement (so the op amp can be operated in a linear region).   

    The purposes of the reference electrode is to take a measurement w.r.t. a reference potential. GND may be a reference potential, but it is not a good one since the conductivity of electrolyte varies from pH setup to setup. A reference potential is more stable, if it is placed next to or inside of pH probe (it will remove the level of IR drop in the electrolyte from different size of pH setup configuration). I have explained indirectly in the previous reply. 

    Please follow the application note, if you feel more comfortable with the pH measurement. 

    Best,

    Raymond

  • Hi Raymond.
    Thanks for your details.

    I will make order to test and come back to report.
    One additional, do you have any comment for PCB layout/placement for Vref and Opamp on board?
    Will you able to review the schematic?
    If yes, i will send you email to your ib.

  • Hi Alone173,

    The integrated LMP91200 sensor is supported by the sensing team, please contact the team for further assistant. 

    Please take a look at the REF35, the specification seems to be good for pH application; and it is on the low cost Vref category. 

    https://www.ti.com/lit/ds/symlink/ref35.pdf?ts=1691537807441&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FREF35

    If this is a discrete pH measurement using LMP7721 or similar PRAMPS op amp, I can be assistant. Please let me know. 

    Best,

    Raymond

  • hi Raymond
    Thanks for your recommend.
    I will try to test and come back with the report.

  • hi Raymond
    Thanks for your recommend.

    Its seem we miss understand.
    I want to replace LMP91200 with op-amp and Vref ICs.
    So i will not use a pre PH interface IC.
    With your recommend, will i use with vref REF35 and 2x LPV821 op-amp for PH circuit?

  • Hi Alone173,

    Are you continuing to use LMP91200, but you want to low cost in some of op amp and Vref design?

    => yes, you're right.

    This is my initial understanding. 

    To replace LMP91200 with LMP7721 or other similar op amp does not lower your overall cost. You will require to have low Ib precision op amp, and you do not have other options. Please advise.

    Best,

    Raymond 

  • Hi Raymond
    My purpose to replace LMPxx to opamp and Vref.
    LMPxx ICs is more expensive than opamp.

    Could you give a parameters for opamp requirements?

  • Hi Alone173,

    Could you give a parameters for opamp requirements?

    These are the precision op amp that may be suited for pH application. Do you know the impedance of your pH probe?

    https://www.ti.com/amplifier-circuit/op-amps/precision/products.html?login-check=true#480=1&3247max=0.02%3B1&sort=3247max;asc&

    Previously, you mentioned that you want us to review the pH design schematic. If the design is a discrete pH op amp design, you may send it to me. I can take a look. LMP7721 is good for the pH application, OPA928 is better, but it is higher in price as well.  

    Best,

    Raymond