• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 63 subscribers
  • Views 1942 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

BQ25504 - suitable for low & high output current config?

FnGas
Prodigy 120 points
Other Parts Discussed in Thread: BQ25504, BQ25570, TPS62737

Good evening.

I am designing a solar power supply that needs to generate outputs for two different loads over a 24 hour period. The MCU load @3.3V is approx as follows:

  • 8mA continuous
  • 80mA for 20mS 720 times a day
  • 250mA for 15mS 720 times a day

The main load current @3.3V requirements are quite high at approx:

  • 16A for 5mS 48 times a day
  • 7A for 50mS 48 times a day

Detailed load analysis reveals the circuit requires ~0.2Ah per day to keep the system operating so for my preliminary design I selected the BQ25504 coupled with a MS-M5050 solar cell and 1100mAh LiFePo4 battery.

The spec sheet of the BQ25504 states the load is to be connected to VSTOR but I'm not sure the IC could handle the peak current requirements of the main load let alone the MCU.

So my questions are:

  • Is the solar cell a good match for the BQ25504? Is there a list of recommended solar cells for this IC?
  • Can I connect both the MCU and main load directly to VBAT? What are the pros/cons of this config? 
  • Can I connect the MCU to VSTOR and main load directly to VBAT? What are the pros/cons of this config? 
  • Can the BQ25504 handle the current requirements of the MCU if connected to VSTOR? 
  • What sort of loads can be successfully connected to VSTOR?
  • Is there a better MPPT IC choice given the unusual current requirements?

Thanks in advance,

Franc

  • 0
    TI__Guru**** 166336 points
    • Is the solar cell a good match for the BQ25504? Is there a list of recommended solar cells for this IC?
      • No. The bq25504 is a boost converter based charger and therefore expects its input voltage to be less than its output voltage.  The MS-5050 output voltage is higher than 3.3V so it is not a good fit for the bq25504.  A panel with open circuit voltage < 3.3V/0.8 = 4.125V (and therefore MPP voltage < 3.3V) is required.  I have attached a spreadsheet that may help you choose the optimal panel for your application.  You will need to complete the items highlighted in red.
    • Can I connect both the MCU and main load directly to VBAT? What are the pros/cons of this config?
      • Yes, in fact you must connect the MAIN load to VBAT because the internal FET between VSTOR and VBAT cannot handle the main load currents. I suggest that you add additional low ESR capacitors in parallel with the battery to help reduce the voltage dip with the main load transients occur.  Otherwise the 16 A times the PCB resistance+battery impedance will cause a large droop in the voltage at VBAT and VSTOR.  If the MCU can operate at a lower voltage then I suggest adding a linear regulator or step down converter to provide power to the MCU to help protect against this droop.
    • Can I connect the MCU to VSTOR and main load directly to VBAT? What are the pros/cons of this config? 
      • Yes.  See question above.
    • Can the BQ25504 handle the current requirements of the MCU if connected to VSTOR? 
      • Yes assuming the solar panel is sized correctly and receives adequate light.  See attached spreasheet.
    • What sort of loads can be successfully connected to VSTOR?
      • As with all boost converters, you have to perform an efficiency balance to determine the maximum output power for a given input voltage: efficiency = Pout/Pin = VSTOR*ISTOR / (VIN*IN).  You can estimate efficiency from the datasheet efficiency curves.  IINavg-max=100mA, VIN = solar panel open circuit voltage *80% and VSTOR is your battery voltage. 
    • Is there a better MPPT IC choice given the unusual current requirements?
      • If you have enough light to keep the battery charged and can size the solar panel appropriately then this solution should work.  If your MCU has I2C capability, then I could recommend a slightly larger battery charger that would require a solar panel with V(MPP) > 4.2V. 

     

    bq25504SolarAppDesign.xlsx
  • 0 in reply to Jeff F
    Prodigy 120 points

    Thanks Jeff, great stuff.

    Does the BQ25504 shut down when there is no power coming into the system and connect VBAT to VSTOR to power the system load (MCU in this case)? Can you recommend a LDO reg for a 250mA @3.3V MCU?

    I guess once the battery is charged and the device is collecting energy from the sun every day all is good, a problem may arise if VBAT is low and the BQ25504 needs to start from scratch to charge the battery, with the MCU load on VSTOR. What is the benefit of connecting the system load to VSTOR? Why not just connect it to VBAT? Maybe if the load is connected to VBAT it could potentially deplete the battery below minimum voltage which isn't good. Would I need a switch on the power input to the MCU to ensure the battery is sufficiently charged before operating the MCU so VSTOR isn't loaded down? 

  • 0 in reply to FnGas
    TI__Guru**** 166336 points

    The FET between VBAT and VSTOR is controlled by the VBAT_UV threshold.  As long as VSTOR is above that threshold, the FET is closed tying VSTOR=VBAT, regardless of the voltage on VIN_DC.  The main boost charger is actually powered from VSTOR. 

    If VSTOR falls below the VSTOR_CHGEN threshold (~1.8V) for any reason (e.g. a 16A load transient), the IC goes into cold start.  You are correct that your loading will likely prevent the IC from exiting cold start.  The reason I suggested connecting the MCU load to VSTOR was to give the VSTOR rail some isolation (via the FET between VBAT and VSTOR) from the main load transient.  Your cold provide that isolation with an external FET driven by VBAT_OK signal as well.

    I suggest you examine the 16A and 7A transients on LiFePO4 output voltage before building anything.  I suspect the transient will collapse the battery's voltage way below 1.8V, due to its own impedance.

    Regarding a linear regulator providing 3.3V, if the output of the bq25504 is 3.3V, then using a linear regulator with 3.3V doesn't really help provide isolation.  I was assuming the MCU would be run at lower than 3.3V. 

  • 0 in reply to Jeff F
    Prodigy 120 points

    The MCU can run on 2.5-3.6V, ideally 3.3V. I could set VBAT_UV to 2.5V however the battery can be discharged down to 2V so this is why I thought I needed a regulator of some sort to maintain at least 2.5V to the MCU. 

    The impedance of the LiFePO4 cell is <25mohm at 1kHz, so I expect this to translate to a 0.4V drop when 16A is drawn. I guess I should adjust VBAT_OK accordingly to ensure it is 0.4V above minimum discharge voltage. 

    Based on your boost converter efficiency balance comment earlier I imagine connecting the MCU to VSTOR, which requires up to 3.3V x 0.25A = 0.825W, would not allow the BQ25504 to climb out of cold start with a 0.4W power input and VBAT<VBAT_UV. As such it seems a FET for the MCU is definitely required so I might as well connect the MCU to VBAT via a regulator and FET. Does this seem fair? Would it still be best to connect the MCU to VSTOR via a regulator and a FET?

  • 0 in reply to FnGas
    TI__Guru**** 166336 points

    The VBAT_UV feature controls the FET between VSTOR and VBAT.  It is designed to prevent battery overdischarge by a system load connected to VSTOR.  As long VSTOR doesn't fall below the VSTOR_CHGEN threshold ~=1.8V, the IC will not enter cold start.  If you are not using this feature, then I suggest shorting VSTOR to VBAT.

    Regarding the impedance, I suggest you also include the resistance of the connectors and board traces when estimating the voltage droop.

    Regarding isolating the MCU power, if VBAT_OV is set to 3.3V but will droop with load then you would need a boost or buck-boost dc/dc converter to maintain the MCU at 3.3V.  Are you planning to set the charge voltage at 3.3V or 3.6V?  If 3.6V, then you might have enough headroom for a linear regulator like TPS78001 family of low Iq linear regulators or TPS62737 low Iq buck converter.   Another option would be the bq25570, which is a bq25504 + 100mA version of TPS62737 in one package.  The linear regulator or buck converter output would need an appropriately sized output capacitor to handle the load transients.