BQ24725A: Battery Charger does not reliably select AC Adapter Power Path with Battery installed in system

For the fault conditions mentioned above, I am guessing I am running into one of the current faults somehow. Is there anyway to tell if one of the High Side / Low Side Over current fault conditions had occurred or maybe in the Input AC Adapter Over current? I don't see a status register anywhere in the part that would show these conditions.

Hi Christian,

For this application, the AC Adapter voltage range is 15V-20V. To accommodate the wide input range, a 3.00 volt shunt reference was added in place of the 100K resistor R49 (R49 is not installed) so that the ACDET pin is clamped at 3.00 volts for any input voltage over about 12V. The DC_IN voltage is 20V and the ACDET voltage is 3.00 volts when the Charger enters the fault condition.

Randy

Hello Randy,

I have a few questions for you for when the fault occurs:

• What is your input current?
• What is your system load?
• Can you measure the voltage on the VCC SRN and SRP pins?
• Does this fault occur at lower system loads?
• Is the adapter able to charge the battery with no system load?

Best Regards,

Christian.

Hi Christian,

Here are answers to your questions.

1) Input current: With no battery installed, the system draws 0.5-1.0A at 20V for the conditions tested above. With the battery installed, the charger will be set to operate at 896mA with system powered ON or 2560mA with system powered OFF. So, when system is powered ON, the system draws about 1-1.5A at 20V for a battery that is not full. With the system powered OFF and the battery charging at its higher rate, the system draws about 1.4A at 20V. 

2) System Load: As mentioned, the system load is about 0.5-1.0A at 20V excluding the battery charger - so about 10-20 watts.

3) VCC, SRN and SRP voltages are somewhat dynamic when I run through a sequence that causes the fault condition. In the case where I have had the system powered OFF, (charger running at 2560mA), I can get the failure to occur by turning the 20V AC Adapter power OFF/ON a few times. The expected state transition (when DC_IN goes from 0-20V) is for the ACOK to become true and the battery charger should switch to AC Power. In the fault condition, the ACOK becomes true, but the charger does not turn ON AC Power although it could be that it tried to switch ON AC Adapter power, failed and then turned OFF the AC Power path. I have looked at the dynamics a bit but do not have a clear picture of what is happening prior to the fault condition. In any case, when the system is in a fault condition, the state of the pins is as follows: VCC=19.3 volts; SRP=10.22 volts and SRN=10.22 volts.

4) The AC Adapter is able to charge the battery when the system is turned OFF. When system power is turned OFF, but the AC Power Adapter is still present, there is a lower power micro controller drawing a small fraction of a watt so that the system load is very small (but not zero) in this condition. The standby Micro Controller is aware of the state of the system and sets the charge rate to the 2560mA or about 30 watts.

I will be looking more at the dynamic behavior of these signals at other when I transition states and cause the fault condition. Any help you can provide to understand what is happening is much appreciated.

Randy Holmberg

Christian,

When I turn DC_IN power ON, there are conditions when ACOK becomes true and DC_IN is only 10V. Sometimes ACOK is nicely delayed from the rising edge of DC_IN, but other times ACOK transitions on the rising edge of DC_IN. When there is no delay, VCC is not 275mvolts greater than SRN and it may go to sleep or something like that. If that is the culprit, then I need to make the shunt regulator that drives the ACDET pin a bit more sophisticated - either longer delay or something like that. Can I change the time delay from when ACDET is valid to when ACOK is asserted? I have bit 15 set to '1' so that the ACOK rising edge deglitch time should be 1.3 seconds. Let me know if you can think of anything to try other than making the ACDET not go valid before DC_IN is compliant?

Randy

So, I think the ACOK timing is not the issue. I have observed a case where the suspect unit was powered up the first time and held off ACOK for about 150msec from the rising edge of DC_IN and the charger still ended up in the fault condition. Attached is a scope trace with DC_IN on channel 1, TP17(output of AC Adapter Mosfet switches) on channel 2 and ACOK on channel 3. In this power up attempt, the TP17 voltage started at the battery voltage of about 11 volts and then jumped up to 20V after ACOK was asserted high - so far so good. However, towards the end of the trace, you can see that the TP17 voltage starts to sag. Sometime shortly after ACDRV was activated, the fault condition occurred and ACDRV switched OFF. Since system loading is very small (Standby power draw only), the voltage decreases very slowly.

Christian,

Some more observations. When the ACDRV turns on, the IOUT has a peak of 0.7volts which would be a peak current of 3.5A. Can you confirm if this would be sufficient to hit the ACOC threshold? If so, is there an easy way to disable this other than clearing ChargeOption() bit[1]?

Randy

Hello Randy,

Randy Holmberg said:

If so, is there an easy way to disable this other than clearing ChargeOption() bit[1]?

Unfortunately there is not an easier way, but I would recommend trying this to see if this resolves the fault you are seeing.

Best Regards,

Christian.