BQ24650: Does not charge properly. Charging current is less then 25mA

Also I noticed if i apply 1.5V to the MppSet pin chagring current increase to around 150mA

Hi Kaushalya Weerakoon,

Charging current could be related to your layout. Could you please check if your sensing routing follows Kelvin sensing(section 11, figure 25)? If not, it could inject error in your charging current. And I also found out you didn't separate IC GND and power GND. Please check if there is any high current flows through AGND to PGND to influence IC reference.

Thanks,

Ann Lien

1. Kelvin sensing : the layout is similer but not 100% the same. both the sence line have the same length and all. dont think this will cause that much of a error ( this is designed for 6.6A and its charing at 150mA )

2. by the lay out i dont think there is high current flow form AGND to PGND ( kind of actually forgot about this )

PGND is is close to the Battery terminal and A gnd connect to it in 2 Places.

any way to see if this is the issue ? ( even so why soo low current ? )

Also any idea why the voltage at MppSet Pin drops ? ( Looks a cap discharge waveform )

Hi Kaushalya Weerakoon,

For Kelvin sensing, could you also check if there is any power supply connect with sensing trace? It may flow high current through sensing trace.

For charging, could you also check STAT1 and STAT2 to verify if it is in charging state?

For Vin drops, could you please check if your PV has enough power to regulate output voltage at 18V?
Do you have our EVM board? Could you also check if this IC is good by swapping EVM IC to your board's IC?

Thanks,
Ann Lien

there are no power lines connect with the sensing paths

Charging : Stat1 is low and stat 2 is high ( Stat1 Led on Start 2 Led Off ) so its in charging mode. if i connect a charged battery ( Li Ion ar 4.1V or so , desinged for LiFePo4 ) leds swap

Im testing using a Programble power supply set to 20V 6A so it has enough power, ( not sure i can test this with a PPS, that could also be the issue ) tried with 18V too but same results.

no i dont have a EVM but i have few boards and all of them works the same way


Could using Agnd and Pgnd same cause this ?

Hi Kaushalya Weerakoon,

So you use bq24650 to charge the battery which's VBAT is lower than 3.5V, it means it is not in CV mode.
Please also check the voltage between SRP/SRN to make sure there is no error in your layout.

Since bq24650 is not like bq24610, which has ISET(charging current setting) need to connect to AGND, it is not directly related to ISET AGND routing. Did you put Vref decoupling cap close to charger IC? I want to know if reference is stable since SRP/SRN refers to IC reference.

Thanks,
Ann Lien

C306 is Bit far away from the IC but close to all the Analog Ground pins

Voltage accroess the SRP and SRN is few mv average with some spikes ( ringing )

Just out of curiosity can we even test this with a const voltage Power supply ? or do we really need a PV panel, or can we test this with a simulated VI characteristic of PV panel ?

Hi Kaushalya Weerakoon,

Could you please send your layout to ann_lien@ti.com? I want to double check if there is any error in your design.

VMPP is a VINDPM loop. If Vin is below the setting value, it will decrease charging current to maintain Vin at your setting value. If you use Vin higher than 18V(you use 20V), it should charge battery. Constant voltage power supply could be used as Vin.

Thanks,
Ann Lien

hi, I sent you the layout Please take a look

i really dont think layout could cause this much issues

Hi Kaushalya Weerakoon,

Let's discuss in mail to speed up our support, thank you!

Thanks,

Ann Lien