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.

TPS65072 unable to detect and charge the drain out battery.

Other Parts Discussed in Thread: TPS65072

We are running into battery detection problem.

When the device is left on the shelf for long time, and battery is completely drain out, the low voltage protection circuit will cut out the cell.

When AC is plugged in, TPS65072 found that no battery is available. And then keep asserting pulse to detect the battery. But the battery is in protection mode and need certain amount of current to recover before the TPS65072 could detect and charge it. So we are running into this dead loop. TPS65072 could not detect and charge the battery and battery remain in the low voltage protection mode.

Another interesting finding is that if the dead battery is inserted before the DC, TPS65072 could recover the dead battery. The reason is that the first pulse voltage is higher than rest of the pulses. 

If the dead battery is inserted after the DC, and missed the first detection pulse, TPS65072 could never detect and charge the battery.

What is the typical solution for the complete drain out battery? Is there a way to walk around this?

  • Hi Hongtao, 

    First, the device checks for a short-circuit on the BAT pin: IBAT(SC) is turned on till the voltage on the BAT pin rises above VBAT(SC). If conditions are safe, it proceeds to charge the battery. 

    Is this the step that you are referring to? You can check by disabling the battery charger then re-enabling it by writing a '0'  then a '1' to bit 0 of register x04. Otherwise, the charger will conduct its battery detect test. This test pulls current from the battery and monitors the battery voltage. For if it drops below VLOWV, then the device determines that the battery has been removed. 

    If your battery is already below VLOWV because, it has drained below it, then your system will always fail this test. The device will continue to test for battery detection. Thus, I believe disabling and re-enabling might help. 

  • It is a little bit challenge in our situation. We are driving the whole system from Vbat. The reason for that is that Vsysout is not regulated.  Now the problem occurs when the battery is dead and there is not Vbat. System could not boot when the charger is trying to detect the battery. Since system could not boot, processor could not talk to the TPS64072 through I2C. So toggling the register 0x04 is not an option.

    Question 1:  The TPS65072 battery detection is not following the specification.  During the 100mA pulse (Iprechg), the dead battery voltage does NOT rise above Vrch (4.1V), but the TPS65072 continue the detection cycle. Has this detection mechanism being tested with a over discharged battery?

    Here is the statement from spec "If the BAT pin voltage falls below VLOWV in the battery detection test, it indicates that the battery has been

    removed. The device then checks for battery insertion: it turns on FET Q2 and sources IPRECHG out of the BAT
    pin for duration tDET. If the voltage does not rise above VRCH, it indicates that a battery has been inserted, and a
    new charge cycle can begin. If, however, the voltage does rise above VRCH, it is possible that a fully charged
    battery has been inserted. To check for this, IBAT(DET) is pulled from the battery for tDET: if the voltage falls below
    VLOWV, a battery is not present. The device keeps looking for the presence of a battery.

    "

    Question 2:  TPS65072 has different behavior the first time AC is plugged in. If the dead battery inserted before the AC. The TPS65072 could detect and charge the battery. But the TPS65072 is doing Fast charge instead of pre-charge. If the dead battery is inserted after the AC, then the TPS65072 is doing the battery detection forever, describe in Question 1.  Based on my experiment, there are two issues. One the battery internal impedance, one is the TPS65072 Vbat detection method.

    The battery has internal impedance, about 150mOHM. And impedance is increasing when the battery cell decaying. The charging current is high (1A), the voltage drop over the impedance is about 0.15~0.2V.  A 2.7V battery cell + 0.2V voltage drop over the impedance, will show 2.9V on the TPS65072 pin. And TPS65072 will detect the battery voltage as 2.9V and start the fast charge instead of the pre-charge.

    TPS65072 has a strong current on Vbat when AC is plugged in. This first pulse will somehow fool the TPS65072 battery detection circuit that battery voltage is above 2.9V.  TPS65072 will charge the battery with fast charging current despite of the battery voltage is below 2.9V. There is no description in the specification what the first pulse is doing. Is it defined behavior or it is just a power on glitch? How we can avoid the high current during the power on so that TPS65072 could charge the dead battery with pre-charge current?

    Too duplicate this behaviour:

    1. Measure the Vbat waveform when the AC is plugged in, you will notice that the first pulse voltage is higher than rest of the battery detection pulse

    2. Insert a "real" drain out battery, the detection will go forever.

    3. Insert a 2.7V battery, and then plugged in AC, PMIC will charge the battery with Fast charge current rather than pre-charge. 

    4. Insert AC first and then insert 2.7V battery, the PMIC will charge the battery with pre-charge current

    How can we work around this issue?

    Regards,

    Hongtao

  • Hongtao, 

    My comments are in RED.

    Hongtao Yan said:

    It is a little bit challenge in our situation. We are driving the whole system from Vbat. The reason for that is that Vsysout is not regulated.  Now the problem occurs when the battery is dead and there is not Vbat. System could not boot when the charger is trying to detect the battery. Since system could not boot, processor could not talk to the TPS64072 through I2C. So toggling the register 0x04 is not an option.

    This can be done with any I2C master like TI's I2C to GPIO box and GUI for TPS65072. 

    Question 1:  The TPS65072 battery detection is not following the specification.  During the 100mA pulse (Iprechg), the dead battery voltage does NOT rise above Vrch (4.1V), but the TPS65072 continue the detection cycle. Has this detection mechanism being tested with a over discharged battery?\

    Here is the statement from spec "If the BAT pin voltage falls below VLOWV in the battery detection test, it indicates that the battery has been

    removed. The device then checks for battery insertion: it turns on FET Q2 and sources IPRECHG out of the BAT
    pin for duration tDET. If the voltage does not rise above VRCH, it indicates that a battery has been inserted, and a
    new charge cycle can begin. If, however, the voltage does rise above VRCH, it is possible that a fully charged
    battery has been inserted. To check for this, IBAT(DET) is pulled from the battery for tDET: if the voltage falls below
    VLOWV, a battery is not present. The device keeps looking for the presence of a battery.

    "

    It sounds like the precharge current x tDET is not enough to charge your dead battery. Restarting the sequence might fix this for it starts with short circuit detection then follows with a long prechg cycle. [tDET = 250ms]

    Question 2:  TPS65072 has different behavior the first time AC is plugged in. If the dead battery inserted before the AC. The TPS65072 could detect and charge the battery. But the TPS65072 is doing Fast charge instead of pre-charge. If the dead battery is inserted after the AC, then the TPS65072 is doing the battery detection forever, describe in Question 1.  Based on my experiment, there are two issues. One the battery internal impedance, one is the TPS65072 Vbat detection method.

    Prechg should be not skipped when AC is plugged in after battery but, the prechg cycle is allowed to be long enough to charge the battery to VLOWV. This is the difference between a battery detect and charging. 

    The battery has internal impedance, about 150mOHM. And impedance is increasing when the battery cell decaying. The charging current is high (1A), the voltage drop over the impedance is about 0.15~0.2V.  A 2.7V battery cell + 0.2V voltage drop over the impedance, will show 2.9V on the TPS65072 pin. And TPS65072 will detect the battery voltage as 2.9V and start the fast charge instead of the pre-charge.

    TPS65072 has a strong current on Vbat when AC is plugged in. This first pulse will somehow fool the TPS65072 battery detection circuit that battery voltage is above 2.9V.  TPS65072 will charge the battery with fast charging current despite of the battery voltage is below 2.9V. There is no description in the specification what the first pulse is doing. Is it defined behavior or it is just a power on glitch? How we can avoid the high current during the power on so that TPS65072 could charge the dead battery with pre-charge current?

    Too duplicate this behaviour:

    1. Measure the Vbat waveform when the AC is plugged in, you will notice that the first pulse voltage is higher than rest of the battery detection pulse

    2. Insert a "real" drain out battery, the detection will go forever.

    3. Insert a 2.7V battery, and then plugged in AC, PMIC will charge the battery with Fast charge current rather than pre-charge. 

    4. Insert AC first and then insert 2.7V battery, the PMIC will charge the battery with pre-charge current

    How can we work around this issue?

    Regards,

    Hongtao

  • Question 1:  The TPS65072 battery detection is not following the specification.  During the 100mA pulse (Iprechg), the dead battery voltage does NOT rise above Vrch (4.1V), but the TPS65072 continue the detection cycle. Has this detection mechanism being tested with a over discharged battery?\

    Here is the statement from spec "If the BAT pin voltage falls below VLOWV in the battery detection test, it indicates that the battery has been

    removed. The device then checks for battery insertion: it turns on FET Q2 and sources IPRECHG out of the BAT
    pin for duration tDET. If the voltage does not rise above VRCH, it indicates that a battery has been inserted, and a
    new charge cycle can begin. If, however, the voltage does rise above VRCH, it is possible that a fully charged
    battery has been inserted. To check for this, IBAT(DET) is pulled from the battery for tDET: if the voltage falls below
    VLOWV, a battery is not present. The device keeps looking for the presence of a battery.

    "

    [TI Reply ]It sounds like the precharge current x tDET is not enough to charge your dead battery. Restarting the sequence might fix this for it starts with short circuit detection then follows with a long prechg cycle. [tDET = 250ms]

    [YHT]: According to spec " If the voltage does not rise above VRCH, it indicates that a battery has been inserted", this is not happening.  When there is not battery, the detection pulse could go up to 4.2V for tDET. When a dead battery is inserted, the detection pulse could only go up to 3.3V for tDET. According to this spec, the TPS65072 should be able to detect the battery and start the pre-charge. But it is not happening. Why?

    Because the PMIC missed the detection during the tDET, it goes to next step to sink 10mA from Vbat. Due to the protection circuit, the dead battery could not provide the current. PMIC failed to detect the battery again.

    So why the PMIC fail to detect the battery during tDET even the voltage is only 3.3V? Should TI update the spec or I am misunderstood the statement?

    Question 2:  TPS65072 has different behavior the first time AC is plugged in. If the dead battery inserted before the AC. The TPS65072 could detect and charge the battery. But the TPS65072 is doing Fast charge instead of pre-charge. If the dead battery is inserted after the AC, then the TPS65072 is doing the battery detection forever, describe in Question 1.  Based on my experiment, there are two issues. One the battery internal impedance, one is the TPS65072 Vbat detection method.

    [TI Reply]Prechg should be not skipped when AC is plugged in after battery but, the prechg cycle is allowed to be long enough to charge the battery to VLOWV. This is the difference between a battery detect and charging. 

    [YHT] I don't think this is the case. 

       a. If you check Vbat pin during the power on, you will notice that first pulse is higher than rest of the pulse. Why?

       b. If the PMIC start with long pre-charge, then it will continue the pre-charge. Because the 100mA pre-charge will only generate 0.06V voltage drop on the batter if the battery impedance is 400mOHM. This will not affect the battery voltage detection. Unfortunately, this is not what we are seeing. The actual test shows that PMIC start with fast charge and then continue the fast charge. You can easly duplicate this behavior using TPS65072 EVM with a 2.7V battery.  What is the difference during the power up vs regular battery detection?

       

    Question 3: What is TI recommendation to deal with the over discharge battery?

  • Hi  Hongtao,

    Has this problem solved or not ?

    We have the similar problem in our design.

    Thanks,

    Jacky

  • No we did not resolve the fundamental issue.

    But we are able to work around it by adjust the power so that battery is connected before DC.

     There is side effect of this method is that the battery is fast charge regardless the battery status.