BQ25713: Zeroing the charge current by hardware

Elder Costa17

Part Number: BQ25713
Other Parts Discussed in Thread: BQ34110,

Battery chemistry: NiMH.

I have the BQ23713 working as Buck/Boost converter for loads and battery charger/controller. A BQ34110 is used as well. Charging/discharging is controlled by firmware using mainly data from the 34110.

For safety reasons, I want to add some hardware mechanism to avoid battery overtemperature to avoid the scenario where the firmware does not terminate fast charging due to some bug. Before somebody asks, I am using the watchdog of BQ25713, but I want to avoid that some condition where the firmware does not detect charging termination and keeps writing to ChargeCurrent() register, causing the charge overload.

I connected one of the ALERT outputs of 34110 to CELL_BATPRESZ pin. I expected that whenever the ALERT output pulled this pin to zero, the ChargeCurrent() register would be forced to zero. However, I noticed that, if I write to ChargeCurrent(), it sets the charging current to a non zero value and it is only zeroed if there is a negative transition on the CELL_BATPRESZ pin.again. So a misbehaved software could inadvertently force the overload. I would like to make the charging system as bullet proof as possible.

So, is there an alternative way to zero the charging current that could be more bullet proof?

Thank you in advance.

over 2 years ago

+1 Khalid Bairuti over 2 years ago

TI__Expert 4780 points

Hello Elder,

I was thinking to suggest setting ILIM_HIZ to LOW to turn the device into HI_Z mode, as I was searching E2E for other examples I came across this question you posted a while back. Have you tried this solution? I don't know of another way without register writes that would work, other than CELL_BATPRESZ (which as you mentioned only zero's the charging current on a H to L transition) or HIZ mode.

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/762648/bq25713-using-ilim_hiz-to-terminate-charging?tisearch=e2e-sitesearch&keymatch=ilim_hiz#

Thanks,

Khalid

0 Elder Costa17 over 2 years ago in reply to Khalid Bairuti

Expert 2825 points

Hello, Khalid, thank you for your comment.

I had completely forgotten that post. I should have digged deeper into the forum or tried different search terms. Thank you for bringing it back to my attention.

PTM would be an option, yes, but I think it also has some down sides, such as the battery having to supply the load until its voltage drops below VSYS plus the P-channel MOSFET diode. But considering the fail safe purpose, this would be acceptable. The partial dependency on software (for configuration) does not seem critical to me. It may be a little more robust than using CELL_BATPRESZ when it comes to software misbehavior, but in the end either has pros and cons.

Your answer is good in both ways, bring an alternative I had completely forgotten and indicates that other than PTM, my proposed approach is viable.

Thank you very much.

Elder.

0 Elder Costa17 over 2 years ago in reply to Khalid Bairuti

Expert 2825 points

One potential issue I see with PTM is I will be connecting the battery to the output of the input power supply (VBUS side). I suppose that could be a problem.

0 Elder Costa17 over 2 years ago in reply to Khalid Bairuti

Expert 2825 points

Khalid.

FWIW, I did a quick test with HIZ mode, it works fine and it is definetely better in terms of robustness compared with my OP idea.

Once again, thank you for your support.

BR,

Elder.

0 Khalid Bairuti over 2 years ago in reply to Elder Costa17

TI__Expert 4780 points

Hi Elder,

Good to know, glad that helped! Thanks for the update.

Khalid

Power management

Power management forum

BQ25713: Zeroing the charge current by hardware