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.

BQ25125: BQ25125YFPR Charger Fault

Part Number: BQ25125

I am seeing a fault in the status register when we have a low input voltage on the charger input. Please note that this voltage is not so low as to trip the undervoltage state. When I look at the fault register there are no faults, which per the datasheet would indicate an ILIM fault. However, I have monitored the current with an ammeter and a scope w/ current sense resistor and am not seeing any current value that would trip ILIM (400mA)

I am suspect to the fact that the input voltage at which the fault occurs is dependent on the SOC of the battery. Further, it seems to correlate with the Vsleep threshold. For example, if I have a battery 3.706V, the definition of Vsleep dictates that it would enter the sleep state between a VIN 3.771 and 3.826. The fault occurred at a VIN of 3.78V.

My suspicion is that we are entering this Vslp state. I am curious as to whether entering Vslp would result in a fault in the charger status register

  • Hello Nicholas,

    Based on your indications, the reported fault and pause in charging is most likely due to entering sleep mode. You can confirm it by raising the VIN above Vsleep + Vbat and confirming that the fault is no longer reported. I see that the datasheet seems to indicate that this is necessarily an ILIM fault but I don't  believe that's the case in this situation.

    Best Regards,

    Juan Ospina

  • Hi Juan,

    That is what I am thinking as well. I have seen the device exit the fault state once it is taken out of the sleep state. I mainly want to confirm that entering sleep state could cause a fault state on the IC.

    I ask as the next issue to tackle is what is causing the problems with the input source leading to these fault states. To address that effectively, I want to ensure the root cause is in fact this sleep state and not something I am missing with ILIM.

  • Hi Nicholas,

    That's understandable, sorry for the confusion. Please let me know if there is anything further I can help with this issue.

    Best Regards,

    Juan Ospina

  • Hi Juan,

    Yes, I am looking for some additional confirmation from TI that this is a realistic reason for the fault flag. Is there any sort of additional documentation where this would be logged?

  • Hi Nicholas,

    ILIM should only come into play if you are drawing close to your ILIM current from your power source. The ILIM is defined by your selected external ILIM resistor before any I2C interaction. Once I2C interaction occurs, default mode is exited and ILIM is now defined by the ILIM register value. By default this register value is 50 mA. It must be written to to increase it to whatever you might need up to 400 mA.

    The Sleep Mode section of the datasheet indicates that it will result in FAULT bits being reported for charging status, though there is no other bit in other registers to indicate that it may be in sleep mode. I believe the line regarding an ILIM fault in the BQ25125 datasheet may not have been necessary since Sleep mode can also cause this status. Based on the fault seemingly dependent on VIN, I believe it is reasonable to assume this is related to Sleep Mode.

    I have confirmed this behavior on my bench as well. Based on your measurements of IIN not having been close to ILIM this is a reasonable cause for the Fault flag. You can confirm on your setup that VPMID tracks battery voltage in this scenario, indicating the battery discharge FET has been enabled as expected in Sleep mode. 

    Best Regards,

    Juan Ospina

  • Hi Juan,

    Thank you so much for your help on this. It is good to see that note in the documentation. I had missed that previously. It is also good to hear you have been successful in reproducing the issue yourself. I will perfrom the testing you mentioned to further confirm behavior.

    Can you advise as to whether there is any way that a processor polling the charger IC would be able to differentiate from a fault being ILIM or sleep state?

  • Hi Nicholas,

    Via I2C alone I don't believe there is a way to distinguish the two but the /PG signal can provide context as it goes high when VIN < VBAT + VSLP but should be low so long as VIN is high enough to provide charge current.

    Best Regards,

    Juan Ospina

  • Hi Juan,

    Thank you for all the help. I apologize for the late reply.. We had one last follow up. I have noticed that when the current is limited on the supply it causes the voltage to droop (as expected). Depending on how much this current is limited, it causes an oscillatory period where the charger turns off and on frequently and faults are logged. However, the device does not toggle the PG pin. My assumption is that it is entering the sleep state, but exiting too quickly to toggle PG? Would you be able to provide any insight into this issue?

  • Hi Nicholas,

    Would you be able to provide a waveform of the input voltage? I'm interested in seeing the frequency of oscillation and the voltages across which it oscillates. Additionally, what current is the supply limited to and what is the charge current and current draw on SYS and PMID? Lastly, what are the logged faults

    Best Regards,

    Juan Ospina

  • Hi Juan,

    This is the AC coupled input waveform to show ripple. This ripple is centered around 4V. Some other things to note:

    Battery Voltage is ~3.765V.

    Oscillation occurred when supply was limited to 40mA

    Oscillation is 35.7Hz, 200mVpp ripple.

    Max current draw I saw was 71mA.

    Fault was detected in status register

    Fault register logged 0x0c, meaning no faults detected

  • Hi Nicholas,

    I'll attempt to recreate this set up on my side and try to find the cause.

    Best Regards,

    Juan Ospina

  • Hi Nicholas,

    I wasn't able to see this when recreating the behavior. For further context, when a power supply is current limited below the fast charge current, the powersupply voltage will droop until it hits the VINDPM threshold. At this point the charger will stop drawing additional power so the powersupply should be stable at this current and voltage.

    Can you re-attempt this test with a different power supply and see if the behavior persists?

    Best Regards,

    Juan Ospina

  • Hi Juan,

    I will retest. Looking at our FW, we do not actually have the VINDPM enabled. Is this something that was enabled during your testing?

  • Hi Nicholas,

    VINDPM was enabled during my testing. This feature is primarily to help with situations such as these where PS current is limited, so this may be useful if this is a likely scenario. I can try my test again as well with VINDPM disabled.

    Best Regards,

    Juan Ospina

  • Hi Juan,

    Have you had any luck with retesting with VINDPM disabled?

  • Hi Nicholas,

    Unfortunately I haven't been able to test this recently. I will do so and provide a response by Monday.

    Best Regards,

    Juan Ospina

  • Hi Nicholas,

    As an updated, I'm unable to recreate the oscillation that you are seeing with VINDPM disabled. I've set fast charge and ILIM to 200mA and 400mA respectively, however when current limiting my power supply at various current limits I'm unable to recreate the oscillation. Have you attempted this test again with other power supplies? What were the conditions of your test in terms of ILIM, fast charge current, input and output voltages, and if the device is in host mode vs default mode.

    Best Regards,

    Juan Ospina

  • Hi Juan,

    I will circle back on your questions

    For our reference, why is VINDPM disabled on startup. I see the datasheet mentions that its for a discharge battery, but I am unsure why that would matter. Is it that the battery voltage would be so low that it would be stuck in VINDPM?

  • Hi Nicholas,

    VINDPM is disabled to allow charging and current to be pulled from a low-voltage source. Mostly useful for a discharged battery because a low-voltage source will only be useful if the battery is sufficiently discharged. If the battery that has high voltage then a low-voltage source may simply enter sleep mode.

    Best Regards,

    Juan Ospina

  • Hi Nicholas,

    Thought I'd check here and see if there were any updates. We were unable to recreate your oscillation and wanted to see if you had the same behavior on other power supplies?

    Let me know if you had further questions about VINDPM.

    Best Regards,

    Anthony Pham