• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 335 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.

BQ25570EVM-206: Low Power Harvester Power Management IC for Hi-Z, Hi-V application

Jason Olson
Prodigy 70 points
Part Number: BQ25570EVM-206
Other Parts Discussed in Thread: BQ25570,

I have a BQ25570 dev board I have been experimenting with: BQ25570EVM-206

The application is a Hi-Z coil that generates 0-130VAC from a magnetic field.  The coil DCR is 350Ohms.  It is connected to a diode (non-synchronous) bridge rectifier, adequately filtered, and fed into a 150VDCmax buck SMPS to output 1.8-5.0VDC, it is set at 3.3VDC.  On the topic of serial connection of multiple control loops; I can go into detail of it's design however the initial questions should not require this info.  This 3.3VDC output is "clean" by most standards when the input power is adequate. The first stage buck has no UVLO and will close or pass the source when under Vreg (1.8VDC-5.0VDC). There is also no boost function but CM and DCM is obviously in play. The "load" is a BQ25570 (via noted TI EVM) with a Li-poly battery on VBAT and VSTOR has been addressed with the datasheet capacitance. There are no other loads on VOUT or what is also sometimes labeled VSYS.    

The mag field for the source coil is mag cycle or RPM based. I am concerned about the boundary conditions and hence the attraction to this type of integrated controller.  The input power is capable of 1W output post the buck SMPS however my issues are not in this region of operation.  The issue is in where the BQ25570 is "asked" (better defined as expected) to cease charging and rely on VBATT for any load needs.  

Below present my questions and issues relative to the BQ25570;

1)  The VBAT output does not appear to maintain a minimum impedance CC regulation or disconnect when the source impedance can not meet the demand or provide a healthy input.  VIN will totally collapse the source. This does not appear to follow the purpose for PV and other Hi-Z applications.

2) In the state defined above, without disconnecting the load, the system is totally unrecoverable even if the source recovers or becomes near ideal.  Disconnecting VBAT remedies the issue.

I would like to know the parameters of the BQ MPP controller with the source.  Is it not compatible with another SMPS or control loop at the input?  Perhaps the VIN "test pulse" time does not align with the input buck's di/dt slaving the capacitance.  I should note the input capacitance ultimately does not matter.  If the input impedance monitoring for the BQ25570 is purely V based, there should not be an issue.  The system as described never recovers as noted above. 

The purpose is to provide a Li-ion/Li-poly charge controller that is compatible with a Hi-Z source that also produces its power from Hi-V, Low-I relative to PVs.  I require pre-stage power management to meet the BQ25570's low input voltage range. The source coil has been tested with a few buck converters to non-intelligently (passive CC limit) charge the same batteries and ultra-cap or super-caps without issue in recovery.  The obvious caveat is that the cell is not allowed to back-feed into the DC shorted source coil through the first buck stage, this is not the case.

Thank you

  • 0
    TI__Guru**** 165886 points

    Hi Jason,

    The BQ25570 MPP circuit allows its boost converter to switch in DCM, pulling up to 285mA peak current as long as the voltage at VIN_DC does not fall below the MPP% x open circuit voltage that is sampled and held on the VREF_SAMP capacitor every 16s.  You can see example operation in d/s figures 26 and 27 on d/s page 26.  If the loaded VIN_DC (in other words not the open circuit voltage) is higher than the VBAT_OV value, a pull down NFET on VIN_DC turns on to pull down the input so that the battery is not over-voltaged. 

    What is the MPP% that you are using? 

    Regards,

    Jeff

  • 0 in reply to Jeff F
    Prodigy 70 points

    Thank you Jeff

    I have run the application at 50%, 70%, and 80% MPP.  I record no differences.   The design is releasing the source every 16seconds.  Here are 2 more observations:

    1) With no load on Vout or VBAT the BQ25570 appears to be unstable and sinks random current from the source.  

    2) With no load on Vout or VBAT the BQ25570 pulls the source SMPS down from 5V.  

    I have experimented with 2 different EVMs and both act the same.   

  • 0 in reply to Jason Olson
    TI__Guru**** 165886 points

    Jason,

    Can you test with a current limited power supply or a regular power supply and series 10-100 resistor?  The supplies should be set slightly higher than the VBAT_OV to prevent the pull down FET from turning on.

    Regards,

    Jeff 

  • 0 in reply to Jeff F
    Prodigy 70 points

    Thank you again Jeff, I am going to tag this resolved.

    I did test the circuit with a linear bench power supply at 1.8VDC-5VDC and a 100OHM series R.   The BQ device appears to behave as intended, including stable clamping of the input voltage.  Inserting the same 100OHM series R into the circuit between the original buck convertor calms the perceived stability issue at room temp but does not fully maintain a fixed input-to-output efficiency as it will still shunt dump current irregularly. Actually with or without the series R.  I am going to proceed to abandon this approach in favor of a CCCV Li-ion battery charge controller with "power path" ....As a low <20mA charge current and add a voltage supervision circuit (with sufficient time constant) to disconnect the charger load from the primary AC-DC stage if it is ever pulled out of regulation.  As of now, I can harvest 400mW at a 5VDC regulated output at room temp and the lowest generator speed the coil would see. 

    "The supplies should be set slightly higher than the VBAT_OV to prevent the pull down FET from turning on."

    I understand as lower.  If TI is listening, a similar device with a buck-boost or just buck stages to the secondary battery and SYS/Vout would be advantageous.  The device in application assumes low voltage, minimum PV cell applications.  However my issues seem to reside purely in the MPPT part of the solution.  

    -Jason

       

  • 0 in reply to Jason Olson
    TI__Guru**** 165886 points

    Jason,

    Sorry this IC didn't work for your application. 

    Thanks for the heads up about the datasheet typo.

    I will communicate your suggestion to our systems and design team.

    Regards,

    Jeff