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.

BQ34Z100-G1: Problem with LOAD MODE and Learning Cycle

Part Number: BQ34Z100-G1
Other Parts Discussed in Thread: BQSTUDIO, BQ34Z110

Good morning, I have a question related to the learning cycle of my fuel gauge. I am using PbA 12V, 7Ah battery. Moreover, I am using the bq34z100G1. I made some configurations since the BqStudio and using the EVM2300. One of the most important configurations are the following ones:

 

Taper Current = 100mA (default)

FC SET% = -1

CC Threshold = 6300 mAh

Design Capacity = 7000 mAh

Design Energy = 14000 mWh

Cell Charge Voltage T1-T2 = 2300mV

Cell Charge Voltage T2-T3 = 2300mV

Cell Charge Voltage T3-T4 = 2300mV

VOLSEL = 1 (external voltage divider)

GNDSEL = 1 (GND of circuit in PACK-)

TEMPS = 1 (external temperature sensor)

Numbers of series cell = 6

Flash Update OK Cell Volt = 1000mV

Load Select = 3

Load Mode = 1

Cell Terminate Voltage = 1750mV

Cell Term V Delta = 100mV

Qmax Cell 0 = 7000mAh

 

After making the configuration I program the CHEMID equal to 808, according to the battery I am using which is a YUASA 12V 7AH NP7-12. FInally, I made the calibration of this learning cycle process as follows:

 

Step1:

I start the process with a full charged battery (100%) and relaxed for an hour after it have been charged . I execute IT_ENABLE and then RESET command since the BqStudio. After sending this commands, RUP_DIS is enabled(red), QEN is enabled(red) too. (QEN remains in this configuration during the whole process)

 

Step2:

After sending IT_ENABLE and RESET. I proceed to discharge the battery to a 1A approximately, what is equivalent to C/7 (where C is the capacity of the battery). Actually, the current is variant and it changes between 900mA and 1100mA during the whole discharging process. When the voltage is high the current is 900mA and when the voltage is low, the current is 110mA. This condition takes place since the 12,8V until the 10.5V approximately. Then, the charging current decrease until 400mA approximately and it remains in this value until the battery’s voltage is 9.3V. The termination voltage is 10.5V However, I maintain this process until 9.3V because of the fact that when I disconnect the battery, the voltage increase considerably. This is a recommendation I got from the specifications of the bq34z110 chip.

 

Step3:

Once the battery’s voltage is 9.3V, the load is disconnected and the battery is leaved in a relaxed state. In other words, the battery is without load, or with the charger connected to to the battery. The battery is leaved in that way, until the OCVTAKEN bit is set (red). In this time, the RUP_DIS bit is disabled(green). It is recommended to leave the battery to relax 5 hours as minimum. However, I just wait until the OCVTAKEN is set to 1, and it takes just 2 hours approximately.

 

 

Step4:

I connect the charger to the battery. Previously, I configure the charger to a charging voltage of 13.8V and limiting the charging current to 1.7A. Due to the fact that I configured the chip with the FC SET% = -1, this FC bit is set (red) when it accomplish the TAPER CURRENT condition, which has a value of 100mA. The discharging process takes 9 hours and when the FC bit is set(red) I disconnect the charger due to this condition.

Step5:

After the FC bit is set to 1(red). I proceed to disconnect the charger and the battery is leaved in a relaxed state until the OCVTAKEN bit is set to 1(red). It happens after 40 minutes approximately. Although it is recommended to leave the battery to relax for 2 hours, I just wait until the OCVTAKEN is set to 1(red). As I am working with a PbA battery, in this stage of the process the maximum error continues in 100% and the Learning Status remains in 4, in other words, these values are not updated.

Step6:

The next step is to connect the load to the battery. The battery is discharged until C/7 what is equivalent to 1A. Although, this value varies between 900mA and 1100mA since the beginning of the discharge until 10.5V. (This has been already explained in the step 2). When the voltage is 9.3V, la current that the load consumes is 400mA. At this point, the load is disconnected from the battery and the battery is leaved to relax.

Step7:

The battery remains in a relaxed state until the OCVTAKEN bit is set(red). It takes 40 minutes since the disconnection of the load. At this point, the maximum error changes to 3% and the Learned Status changes to 5. I understand that when I am using a PbA battery, it is normal that those values get updated during the discharge. In the case of other chemicals, these values should be updated during the charging stage. Although. I am not so sure about it, that is what I supposed, I just have worked with PbA batteries until now.

 

Step8:

I repeat the steps 4 and 7 one more time until the maximum error is 1% and the Learned Status is equal to 6. In both charging processes the FC bit set(red) when the CURRENT TAPER (100mA) condition is accomplished(FC SET% = -1)

 

 

At this point I finalized the learning cycle process. The LOAD MODE parameter was configured for CONSTANT CURRENT. However, my load is a POWER CONSTANT load.

I proposed to develop the learning cycle by changing the LOAD MODE value from 0 to 1(which corresponds to POWER CONSTANT) and doing this, I get the following problem in the Step4: The FC bit never set and I dont know why, the charger current in this point is less than taper current and the SOC is 100%. I don’t know if FC bit set is a necessary condition. I have read some publications in which the learning cycle process procedure of PbA batteries doesn’t mention the FC bit during the charging stage. The procedure just looks for charging to 100% of SOC, and then it waits until the OCVTAKEN is set. I would like you to help me to solve this doubts. Thanks.

 

My data Memory.

DataMemory.zip

 

  • Hi E-Idea Tech,

    Thank you for the very thorough post. If FC SET %=-1 then the FC bit will set when current taper conditions have been met. Otherwise, the FC bit will set to whatever FC SET %= (default 100%).

    Please use the constant current setting for the initial learning cycle so the device may learn the cells. After the device reports an update status of 06, please modify your load settings to your preferred load settings for gauging.

    When learning with lead acid, the device will update on the discharge only, so the steps are as follows:

    1. Enable IT, Update status should now be 0x04
    2. Charge to full and allow the battery to fully rest
    3. Discharge at a constant current until true termination voltage is reached (found in the datasheet of your cell)
    4. Allow the cell to fully rest
    5. Update status will update to 0x05
    6. Repeat steps 2-4. Update status should now be 0x06

    Sincerely,
    Bryan Kahler

  • Hi Bryan Kahler,

    Thanks for your help . I have some questions related to your answer:

    It is ok to make the learning cycle process with CONSTANT CURRENT but using a CONSTANT POWER load in my final application? Does it affect the measurements of the device? In other words, will it continue to be accurate in the measurement of parameters related to time, SOC,SOH,etc.

    About the charging cycle, I am going to mention your steps to clarify it better.

    1. Enable IT, Update status should now be 0x04

    a) In this step, should the battery be charged, discharged or the charging level that is going to be used in the process is not important ?
    b) Should the battery be relaxed before the learning cycle process begins?
    c) Should the battery be a new one for using it in a learning cycle process? Is it necessary that or not?
    d) Should I send a RESET command after sending the IT_ENABLE command? Is it necessary or not?

    2. Charge to full and allow the battery to fully rest

    a) When the battery is charged, should I wait until the FC bit is set or should I just wait for a 100% of SOC? In other words, is it the FC bit
    important for the process?
    b) When disconnecting the battery and leaving it relaxed, is it enough to wait until the OCVTAKEN sets(red) that occurs in about 40
    minutes. Should I wait for 2 hours according to the steps that the manual of the chip recommends or should I wait the time that the
    manufacturer specifies for the battery which is about 24 hours?. In the last 2 cases, the OCVTAKEN bit would be set(red). I don’t know
    if that can generate any mistake.

    3. Discharge at a constant current until true termination voltage is reached (found in the datasheet of your cell)

    a) It is important that the discharging current is exactly C/7 (1A in this case) or can it vary between 900mA and 1100mA as it is my case?
    b) The load I am using for my final application varies between 1.5A and 2.5A. If I make this learning cycle process as if I were working
    with 1A, is it possible to make mistakes or get errors in the measurements? Or, should I use a discharging current similar to the
    current of my final application?
    c) The termination voltage is 10.5V in my case. When I disconnect the load from the battery the voltage increases until 11.5V
    approximately. Is it normal? Can it affect the learning cycle process?

    5. Update status will update to 0x05
    6. Repeat steps 2-4. Update status should now be 0x06

    Sincerely,
    E-Idea Tech
  • Hi E-Idea Tech,

    The answers are inline below:

    >>It is ok to make the learning cycle process with CONSTANT CURRENT but using a CONSTANT POWER load in my final application? Does it affect the measurements of the device? In other words, will it continue to be accurate in the measurement of parameters related to time, SOC,SOH,etc.

    Yes, Please use constant current for the learning cycle and then constant power for your final application, if constant power is the load mode of choice for your application.

    About the charging cycle, I am going to mention your steps to clarify it better.

    >>1. Enable IT, Update status should now be 0x04

    >>a) In this step, should the battery be charged, discharged or the charging level that is going to be used in the process is not important ?

    This is prior to charging.

    >>b) Should the battery be relaxed before the learning cycle process begins?
    Yes, after charging to FC, allow the battery to fully relax before discharge. If there is a large temperature differential between where the battery was stored and the room temperature test environment, allow up to 2 hours for the temperature of the cell to reach equilibrium prior to testing.

    >>c) Should the battery be a new one for using it in a learning cycle process? Is it necessary that or not?
    It is preferred if new cells are being used in production

    >>d) Should I send a RESET command after sending the IT_ENABLE command? Is it necessary or not?
    You may, but it is not necessary

    >>2. Charge to full and allow the battery to fully rest

    a>>) When the battery is charged, should I wait until the FC bit is set or should I just wait for a 100% of SOC? In other words, is it the FC bit
    important for the process?
    Yes, allow the FC bit to set.

    >>b) When disconnecting the battery and leaving it relaxed, is it enough to wait until the OCVTAKEN sets(red) that occurs in about 40
    minutes. Should I wait for 2 hours according to the steps that the manual of the chip recommends or should I wait the time that the
    manufacturer specifies for the battery which is about 24 hours?. In the last 2 cases, the OCVTAKEN bit would be set(red). I don’t know
    if that can generate any mistake.
    Please wait until the cell is fully relaxed. This may be up to 5 hours with a lead acid cell, or possibly more according to the manufacturer.

    >>3. Discharge at a constant current until true termination voltage is reached (found in the datasheet of your cell)

    >>a) It is important that the discharging current is exactly C/7 (1A in this case) or can it vary between 900mA and 1100mA as it is my case?
    Please perform the learning cycle with discharge rate of C/10. It is important that the cell has a large DOD drop for the learning cycle. The higher the rate, the higher the rebound of the cell when resting after discharge.

    >>b) The load I am using for my final application varies between 1.5A and 2.5A. If I make this learning cycle process as if I were working
    with 1A, is it possible to make mistakes or get errors in the measurements? Or, should I use a discharging current similar to the
    current of my final application?
    The rate is for the device to learn the cell. After the pack is learned, you may modify the discharge rate to that of your application when performing SOC accuracy testing.

    >>c) The termination voltage is 10.5V in my case. When I disconnect the load from the battery the voltage increases until 11.5V
    approximately. Is it normal? Can it affect the learning cycle process?
    Yes this is normal and yes is can affect the learning cycle process. To mitigate this, we use a lower rate and ensure full discharge of the cell.

    >>5. Update status will update to 0x05
    >>6. Repeat steps 2-4. Update status should now be 0x06

    You may also temporarily adjust the Cell BL Set Volt Threshold so as to not affect a full discharge.

    Sincerely,
    Bryan Kahler
  • Hi Bryan Kahler,

    Thank you very much for the information, I hope you can help me in another thread.

    Sincerely,
    E-Idea Tech