BQ25756: PG' is active, Stat1 is on, Stat2 is off, but no power being produced

Hi Sean,

Sean Solberg said:

What situations could there possibly be where PG' is on, Stat1 is on and Stat2 is off but yet the BQ25756 is not producing ouput power?

I typically see this when there is a charging fault that the converter cannot see. I believe what happens here is that the charger thinks that it is charging while in reality it is not. This means that the device thinks it is and allowing power transfer switching while there is something that is preventing it from switching.

To know more about what is occurring, I recommend looking at the power stage and the control for the power stage. This would be SWx, LODRVx, HIDRVx, BTSTx, and DRV_SUP.

In this situation, the device would be reading a battery short and could be in precharge mode, which has reduced current. This usually results in the capacitors still being charged, but at a slower rate, and then reaching termination voltage before oscillating between charge and recharge.

Best regards,
Michael

Thanks for the info.   I will be scoping the signals today to see what's happening there.    Here is a video of the behavior. 

The core of the issue is that sometimes it produces power as expected, and sometimes it does not even though we have PG' and Stat1. 

https://www.youtube.com/watch?v=DieLt2ZUFag

..Sean

Hi Sean,

That is interesting behavior. I have not seen that on the EVM, and I have run the same test where I lower the input voltage below the ACUV and above the value again. I did not see this behavior.

When the STAT1 and STAT2 lights start flickering, this might be right at the border of ACUV where drawing current puts the device below ACUV, but no current has the voltage just above ACUV.

Let me know what the behavior is on the switching waveforms. Do you have a schematic you can share? Do you have I2C access?

Best,
Michael

I can confirm that when the BA25756 is not producing power and the input is at about 18V with PG' on and Stat1 on, there are no fault codes reported from the IC via I2C. 

I may have possibly figured it out.   I think what is happening is that the charger begins to produce output power at 9Volts based on the resistors we have on Vacuv.   This is correct.  Then, the CPU fires up and begins talking to the charger IC over I2C and it first gives it our set of I2C based settings.    Even though we have MPPT turned off in those settings, we are setting the VReg_DPM_ACC register to 19 volts.  As soon as that happens, the charger stops charging, but leaves the PG signal on.  And even though it's actually not charging due to being in Input Voltage Regulation, the Stat1 pin is still ON.     As soon as this happens, the CPU that runs our board shuts off because the charger IC shuts it's output off, and then at that point it will never recover.    I will keep investigating. 

Also, the documentation doesn't quite describe this situation accurately as Table 8-6 does state that STAT1 being ON means that "Charge in progress", which is not quite the case here.   

The documentation also states: "When VACUV falls and reaches VACUV_DPM, the device enters input voltage regulation, thereby reducing the charge current. VACUV continues falling below VREF_ACUV, the device automatically stops the converter and the PG' pin pulls high."     At first when I read this, it seems to say that either of the reasons to stop the converter would pull the PG pin high, but technically it's worded in such a way that when it's under Input Voltage Regulation it's just supposed to reduce the charge current, but the charge current is already so low, that it really means it's going to go OFF, but still keep PG' on.

Anyway, I think I'm figuring out the situation in that Input Voltage Regulation doesn't really affect PG' but it does actually turn the power converter switching off.