Planning to use 3.7V non-Rechargable battery with npm2100 with an additional diode?

I am planning to use 3.7V non Rechargable Battery with Diode and Reverse Protection circuit with npm2100, Would like to build a Fuel Guage model for this as this is not available. As per Data sheet. voltage max support is 3.4V and absolute max iw 5.0V.  Let me know if this is an workable solution with npm2100.  Voltage Drop from Diode and Reverse protection circuit will be below 3.4v as per specs. 

Parents
  • Hello Ramakrishna,


    Currently only the following battery models / chemistries are supported with nPM2100 fuel gauge solution.

    AA Alkaline (single or two in series)
    AAA Alkaline (single or two in series)
    Alkaline coin cells
    LiMnO2 based coin cells

    Also note that nPM2100 voltage based Fuel gauge solution cannot be used with battery voltages above 3.2V since ADC is limited to measure VBAT below 3.2V 

    But if the other ratings are within the rated operating conditions of nPM2100 can be used without using fuel gauge. 

    Best regards,
    Ressa



  • Thanks for the response. If we are planning to use 3V LiMnO2 Battery with 3.3Ah, does fuel gauge auto detect this ? or what is alternate solution. we are planning to use LM17500 3V  lithium manganese dioxide A-size spiral cell (Not coin cell). How to plan for Fuel Guage ? Technical specs : https://saft4u.saft.com/fr/download_file/e72b0f83-9ac0-44be-96a2-5bd243d1f8be/English 

  • Hi,

    There is no auto detection mechanism. You select one profile explicitly, either at compile time (Kconfig, e.g. CONFIG_BATTERY_MODEL_LITHIUM_CR2032) or at runtime via a shell/API call that loads one of the built-in profiles. 

    The only LiMnO2 profile available is the CR2032-based coin cell model. Yours has different form factor so the parameters will differ . So CR2032 probably will not be accurate. 


    One option would be to try the CR2032 model as a rough approximation and validate it empirically against real discharge data for your LM17500 cell/load profile. This may be acceptable if you only need coarse SoC bands (OK / low / critical) rather than precise percentage reporting, but Nordic doesn't warrant its accuracy outside the coin-cell format it was characterized on.

    As of now, there  is no self-service characterization tool for primary cells, unlike the rechargeable side.
    Please reach out to regional sales team to evaluate the project in case there is possibility of making custom profile for your desired battery: Contact Us - nordicsemi.com


Reply
  • Hi,

    There is no auto detection mechanism. You select one profile explicitly, either at compile time (Kconfig, e.g. CONFIG_BATTERY_MODEL_LITHIUM_CR2032) or at runtime via a shell/API call that loads one of the built-in profiles. 

    The only LiMnO2 profile available is the CR2032-based coin cell model. Yours has different form factor so the parameters will differ . So CR2032 probably will not be accurate. 


    One option would be to try the CR2032 model as a rough approximation and validate it empirically against real discharge data for your LM17500 cell/load profile. This may be acceptable if you only need coarse SoC bands (OK / low / critical) rather than precise percentage reporting, but Nordic doesn't warrant its accuracy outside the coin-cell format it was characterized on.

    As of now, there  is no self-service characterization tool for primary cells, unlike the rechargeable side.
    Please reach out to regional sales team to evaluate the project in case there is possibility of making custom profile for your desired battery: Contact Us - nordicsemi.com


Children
Related