BQ24171: the issue at disconnecting AC adapter input

Hi Muk,

What is the BTST capacitor size?  Is there a series resistor with that capacitor?  If capacitor size is different that datasheet recommends or there is a >5 ohm series resistor, the IC could be boosting back.

Is there a 499kohm resistor (Rgs) from CMSRC to ACDRV as shown below?  If not, we have seen leakage cause the ACDRV charge pump to operate incorrectly when single cell batteries are used to power AVCC.

Regards,
Jeff

Hi Jeff-san

BTST cap = 0.047uF

Series resistor = 0ohm

Rgs from CMSRC to ACDRV = 510kohm

I think, they are recommended values. How do you think about the reason of this issue?

Muk

Hi Jeff-san,

Customer tried to modify BTST cap to 0.1uF.

Unfortunately, The behavior is not improved. (not see anything changing)

Could you please give me another solution?

Thanks,

Muk

Hi Muk,

Does this only happen at a certain battery voltage, for example, full charge?  

Is there a diode from BAT to AVCC?  If so, try removing it.

If you lower the OVPSET resistors by a factor 10, does the additional loading on input stop the boost back?

Regards,

Jeff

Hi Jeff-san,

>Does this only happen at a certain battery voltage, for example, full charge?  

Please refer the following table.

> Is there a diode from BAT to AVCC?  If so, try removing it.

There is not a diode.

>If you lower the OVPSET resistors by a factor 10, does the additional loading on input stop the boost back?

Customer tried to check by 1/10 OVPSET. However it is not solved..

Muk

Hi Muk,

Regarding the BTST capacitors, what is their voltage rating?  I want to confirm that DC bias derating is not a factor

Can you add the switching node waveform to the present switching node waveforms for C(BTST)= 0.047uF and 0.1uF?  

How low of a resistance do you need to add from input side of FETs to ground to prevent the issue?

Regards,

Jeff

Hi Jeff-san,

>Regarding the BTST capacitors, what is their voltage rating?

  >> 50V

>Can you add the switching node waveform to the present switching node waveforms for C(BTST)= 0.047uF and 0.1uF?  

  >> Please refer the following waveform

>How low of a resistance do you need to add from input side of FETs to ground to prevent the issue?

  >> Customer try to add 12ohm resister. However the oscillation issue was still occurred. Please find the waveform at adding 12ohm below,

My understanding is below,

1. ACFET is ON at decreasing Vin voltage

2. BATFET turn ON due to ACN<SRN+220mV(typ)

3. Input current is stopped

4. AVCC(&ACN) voltage will be increased due to input current change to 0A and Input resistance and capacitors in AC adapter resistance.

    Or Battery voltage is boost back to input cap

    Also, at this timing, it might have reverse current.

5. BATFET turn OFF Due due to ACN>SRN+220mV(typ) + 100mV

   And, Charging start >> Input current will increase

Repeat 2 to 5

Is it correct?

Customer checked this issue with our EVM and DC stabilized power supply for input voltage.

Customer found the similar behavior at input voltage = Battery voltage + around 300mV.

Could you please check it?

Also, Customer is considering about adding reset IC for monitoring input voltage as below,

Does it have any problem?

Thanks

Muk

Hi Muk,

Essentially I think you are correct.  The input FET turn off time is to late/slow relative to the converter refresh pulse.  Also the adapter output capacitor holds up the input voltage just long enough to create a FB loop.  The reset IC might help but I would set it to a higher voltage, for a 12-V adapter, like 9V or to a % of input voltage if 5V adapter used.  Alternatively, if we can remove components or add a simple Schottky diode to speed up the Q1/Q2 turn off, that should work too.  Although not shown in the schematics above, is your customer using CDQ and CGS has shown in the graphic below:

These caps are used to slow down turn on speed to reduce inrush current but also slow down turn off.  If replaced with lower value or removed completely, does that help?  If they add a Schottky across R11, cathode at ACDRV, does that help? 

I will be unable to look at this in the lab until after the US holiday break.

Regards,

Jeff

Hi, Jeff-san,

Customer tried to remove CDQ and CGS. >> Not improved

Customer tried to add a Schottky across R11, cathode at ACDRV. >> Not improved

Customer is considering to adjust the resisters for OVPSET for 9V or 12V adapter. also, they will use reset IC for 5V adapter.

If you have any other concern, please let me know.

Thanks

Muk

Hi Jeff-san,

Reset IC, open drain output,  Vth around 4.5V .

The waveform are below at adding reset IC.

Thank

Muk