BQ25792: BQ25792:Switching Problems

Hi Stony,

Can you provide a register dump?  What are the values in the status, fault, VINDPM and IINDPM registers?

Regards,

Jeff

Stony,

I noticed that /CE pin is pulled high.  This disables charge current into the battery.  Is the schematic correct?

Regards,

Jeff

Hi Jeff, 

First thank you very much for your attention and support.

Below are the registry dump.

Fault flag (0x26- 0x27) : 0x0000 (No Fault)

VINDPM (0x05) : 0XD2 (21.0V)

IINDPM (0x06 -0x07) : 0x00BF (1.91A)

If I am not mistaken, the above data looks quite normal. 

And by setting register 0x12 to 0x30, we manage to disable the PFM mode and so the BQ25792 can always stay at PWM mode. However, there is still severe ringing in the switching waveform as shown in the previous post and the board continue to draw current from the Battery even when VDC is present. 

We have attempted to provide better ground connection by connecting the gnd among the IC, the PMID capacitors and VSys capacitors with an additional AWG#24 wire but that does not help.

Would the ringing problem be related to the choice of the components or the PCB layout ? Below is our PCB layout. The board is a 4 layer board and there are two internal ground plate not shown in the below image. 

Normally we will simply do more trial and test with different components or even PCB layout. However, as there are currently global shortage on this IC (my search reveals that the earlier parts availability is in 2022) and so we really want to plan carefully before we modify the board so as to avoid any damage on the IC. 

As for the query concerning the /CE pin connection that you raised in another post, that design is intentional as we want to change the charging current to 0.3A before enabling the battery charging. We have double checked that this /CE pin has been pulled low by the MCU during the test. 

Thank you again for your support. 

Best Regards,

Stony Kong

Hi Stony,

Thanks.  I will review the registers in more detail and get back to you.  Regarding the ringing, a converter operating in discontinuous mode has ringing similar to your second plot above.  I would expect the pulse to get closer to ground than what your plot shows, however.  Can you check the BTST nodes and make sure they look like the SW nodes but ~5V higher voltage?

Regards,

Jeff

Hi Jeff, 

Really appreciate your effort in studying the registers. 

And below are the waveform at SW1 and BTST1. 

The BTST1 waveform does appear to be a few volts higher than the SW1 waveform. And also revealed from the above photo, it seems that there were missing pulses in the switching output. 

Thanks in advance for your advices. 

Best Regards,

Stony Kong

Hi Stony,

I assume you were bucking with VBUS>VBAT at the time?  If so, then there is problem with either the IC or with BTST1 capacitor and connection to BTST1 in and SW1.  

Regards,

Jeff

Hi Jeff, 

Yes, the input voltage is at 24V and we are using a 4 cell battery at around16.8V and so the circuit is always bucking.  

Following your earlier reply, we have done the below test : 

1. Change the 0.047uF capacitor to ensure that the board is mounted with 0.047uF capacitors.

Result : No improvement seen.

2. Change the 2.2uH inductor to the TDK SPM6530 series (same model and brand as used in TI EV board).

Result : No improvement seen.

3. Change the 0.047uF capacitor to 0.024uF (just in case it can match better with 2.2uH inductor). 

Result : No improvement seen. 

Hence, apparently the possible causes of the problem is left with either the PCB layout or the chip. Do you see any problem on the PCB layout ? 

We do wish to try a new BQ25792 on our board but unfortunately, we no longer have any new chip to do the replacement. 

Best Regards,

Stony Kong

Hi Stony,

I don't see any obvious issue on the PCB.  Can you remove the IC and place on a BQ25792EVM?

Regards,

Jeff

Hi Jeff, 

Thanks for your checking on the PCB layout and glad to know that there is no obvious problem. 

As for your suggestion of testing the chip on the EVM, I am sorry that we actually do not have the BQ25792EVM. And if the problem is likely on the particular chip, probably we would need to put this design on hold until the chip is available again in the market. And at the mean time, we would need to start our plan B to meet our prototype delivery deadline. 

Anyway, do wish to express our appreciation to your kind support to our project. 

Best Regards,

Stony Kong