This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ24800EVM: Battery boost minimun battery voltage for backup

Part Number: BQ24800EVM
Other Parts Discussed in Thread: BQ40Z50, , BQSTUDIO, BQ24800

I'm designing a UPS power supply, with a pack of 2S LiFePO4 cells (7.4V max - 5V min) and bq40z50 as BMS. I need VSYS to be always regulated at 12V and 2A max current on backup.

I bought BQ24800EVM and made some tests with this board, it all works as expected except when I try to enter backup with VBAT below 5.5V, or switch from loaded VSYS at 2A to battery backup with less than this voltage on VBAT. I would like to lower this threshold as I may want to discharge my cells some more, and let the bq40z50 decide when to stop.

My settings are:

- Charge Voltage: 7440mV

- Battery depletion comparator threshold: Failing Threshold 59.19% ~2.486V/Cell    (this means around 4.4V minimun voltage on the battery right?)

- Battery boost regulation voltage: VSysMin + 1.6V

- Minimun system voltage: 10496 mV

I don't know if this could be modified or if it's a restriction of the IC as I don't have the datasheet yet, I'm waiting for approval.

  • Hi Francisco,

    As you have noted, the VBAT_DEPL setting determines when to exit boost mode due to battery depletion.  59.19% is the lowest setting, however, this percentage is a percentage of the Charge Voltage register.

    Therefore, if you want to extend to a lower battery depletion voltage, you can reduce the ChargeVoltage (Register 0x15) setting while you are in boost mode.  You will then need to set it back to the target charge voltage of 7.44 V once the adapter plugs in and you start charging.  It takes a little uC overhead, but with this trick you will be able to adjust to whatever depletion threshold you want. 

    Just note that with such a large boost voltage of 5V to 12V, there will be 2.5 times as much current pulled from the battery as you are pulling from SYS.  This can lead to some very high currents, and with the battery already so far depleted it may be an issue.  The device will let you set the depletion lower using the trick I outlined above, but you may still run into system issues.

    Regards,

    Steve

  • Hi Steve,

    Today I performed some tests with this setup:

    - BQ24800EVM with battery-only backup configuration, VSYS = 12V on backup,  VBAT_DEPL = 59.19% of  Charge Voltage

    -bqStudio 1.3.101 connected through EV2300

    - Power supply on VBAT

    - Voltimeter on VSYS (No load)

    I configured ChargeVoltage and my power supply on VBAT for 7.5V. I got 12V on VSYS until I lower VBAT to 5.4V, when VSYS regulation stops and I get 4.8V, as VBAT discharges through Q2 MOSFET. This is the same as yesterday to ensure my configuration is OK.

    Next I configured ChargeVoltage and my power supply on VBAT for 7.5V again. Then I lowered VBAT to 7V and modified ChargeVoltage to 6.8V. I got the same result as before, whenever I reach 5.4V on VBAT I lose regulation.

    Finally I configured ChargeVoltage and my power supply on VBAT for 7.5V again. Then I lowered VBAT to 6V and modified ChargeVoltage to 6V. I got the same result as before, whenever I reach 5.4V on VBAT I lose regulation.

    Note that I can NEVER select ChargeVoltage lower than 6.5V on bqstudio as the starting point because it resets the ChargeVoltage register to 0 by itself whenever I do that.

    Thanks for the system level recommendation, it's true that higher currents are not good for a near depleted battery, but we monitor battery status with bq40z50 to check its health. We don't plan to discharge the battery that far always, but we want the ability to do it.

  • Also, what is the BQ24800 default I2C Address? Is it modifiable?

  • Hi Francisco,

    The 7-bit address is 0x9.  If you extend this to an 8-bit address by appending a 0 in the (R/W) bit, it is 0x12.  This is a fixed address, it can't be modified.

    Regards,

    Steve

  • Hi Francisco,

    I tried this out in the lab.  I was able to boost to 12V from a 4V supply at BAT (it was a bench power supply, not a battery.) 

    Here were the settings that I used:

    There must be something other than the VBAT_DEPL threshold that is kicking the system out of boost.  Can you try the settings I have above to see if you can replicate boosting from 4V at BAT?  You might want to try a bench power supply instead of battery just to remove the gauge IC and battery behavior as variables. 

    Regards,

    Steve

  • Hi Steve,

    I replicated your test, with same settings, power supply and no BMS (as in all my previous tests), and it worked correctly, losing regulation around 3.2V as expected

    However that's the same configuration I used before and I'm still unable to get it to work with the voltages I need.

    Can you please try a Charging Voltage of 6V and a backup of 12V, lower VBAT voltage from there and tell me when VSYS loses regulation?

    Because with this exact configuration (only changing Charging Voltage) my board loses regulation at 5.4V on VBAT.

  • Hi Francisco,

    I am working from home some days of the week, but I should get into the lab to try this out on either Wednesday or Thursday and I will let you know.

    Regards,

    Steve

  • Hi Francisco,

    Here is a scope capture with those settings.  When the converter stops boosting, the battery voltage has an average value of 3.648V and a minimum of 3.576V.  Please see scope capture below.

    Regards,

    Steve

  • Thanks for the detailed and quick

    I see that your board is working as expected.

    I have a couple of questions:

    - I see that reel BQ24800 come with different marking than the one on te EVM. I guess it should be the same IC version but just to clarify:

    Do you have this same IC marking on the EVM? Do you recommend replacing it if I'm still unable to get your last test performance?

    - Do you start the test with power on VIN, next configuration and finally disable VIN and the board enters into boost or do you just start the test from a powered off board which you aplly power on BAT and then configure boost mode? Does this sequence matter? I'm doing the latter.

    Regards,

    Francisco

  • Hi Francisco,

    The initial EVMs are made with pre-production ICs.  It takes some amount of time for the stock to deplete and the EVMs rebuilt with the final production silicon, but the pre-production silicon should be 99% equivalent.  Sometimes we make a minor adjustment to a trim target or two before production, but generally the pre-production silicon should behave exactly the same as the production ICs. 

    You could try replacing the IC, but I don't think it would make a difference.  Also, the EVM I am using has pre-production IC.

    I have been running the tests with battery only.  I never plug in the adapter, I power on from battery and then start the boost mode.  It should not matter which method you use to enter boost. Once in the boost mode, it should behave the same regardless of which method was used to enter the boost mode. 

    Regards,

    Steve