USB input, LiPo storage, 3.3V output

We do not have a charger that can both charge a battery and provide a regulated output at a voltage that is between the max and min battery voltage.  However, one of the bq2415x or bq2418x families, with a TPS63020 after the SYS rail to provide the 3.3V rail might work for you.  Do you need an IC will dual input, one for USB and one for the wall wart (i.e., AC input)?

By LiPo, did you mean Lithium Polymer or Lithium Iron Phosphate? 

I'm looking to use Lithium Polymer batteries, preferably 1 x 3.7v cell (in response to another question Jeff had).

With the bq2415x to go above 500mA when attached to a wall wort the chip must be told to do so over I²C, right?  Boost mode is only for getting around 200mA back out the VBUS pin, right?  The TPS63020 looks good too.  Plenty of current at 3.3v from 2.5v-5.5v, right?

The TPS63020 needs additional undervoltage protection to prevent discharging a LiPo below the recommended 3.0v, right?  It looks like the bq2415x has undervoltage protection down to 3.05v.  In order to use that protection, where is the system load attached?  If it is on the battery, the charger has no way to prevent undervoltage.

One of my potential designs will run off of USB only.  Another similar project should have the option to charge off of either USB and/or an AC/DC wall wort.

- Brent

The bq2415x parts require I2C communication for full functionality.  Boost mode is to provide current back to the VBUS pin.  If you don't have I2C and/or don't need boost mode, we have other parts that are "stand alone" and use multiple logic inputs to change the charge functions.    You might be able to narrow down the choices at the website below:

http://focus.ti.com/paramsearch/docs/parametricsearch.tsp?family=analog&familyId=411&uiTemplateId=NODE_STRY_PGE_T

Regarding the TPS63020 question, yes, we would need to add a simple supervisory circuit to disable the TPS63020 if the battery is discharged too low. 

If you need dual input (one for USB and one for wall wart), USB communication, up to 2.5A charge current from wall wart but no boost mode for USB OTG, we have a new part family coming out  later this year.  I have attached the first page of the datasheet.  We also have an I2C version.

Hi Jeff,

Can we use BQ24165 for charging Lithium Polymer battery based design?

Thanks

Ranjith

Yes.

Dear Jeff

I would like to ask you about the safety timer fault on BQ24166.

My Question is as below

I checked the safety process in the red color through testing TI's EVM B'D and find out that toggle the CE (High to Low) to clear Safety timer fault, but the CHG is stayed at High impedance(LED Off in the EVM B'D).

Is right operation that the CHG is stayed at high impedance though clear safety timer fault with toggling CE(High to Low)?

=================================== Safety section From the data sheet ==================================================================

SAFETY TIMER At the beginning of charging process, the bq24165/6/7 starts the 6 hour safety timer. This timer is active during the entire charging process. If charging has not terminated before the safety timer expires, the charge cycle is terminated and the battery FET is turned off. To clear the safety timer fault, charging must be resumed by using CE1 and CE2 (bq24165) or CE(bq24166/7) or High Impedance mode or a new charge cycle started by VSUPPLY POR or battery remove and replace.

=================================================================================================================================

Thanks and Best

Yes.  The /CHG output is independent of the safety timer.  The /CHG only goes low for NEW charge cycle.  A new charge cycle is started by VSUPPLY POR or battery remove and replace.   If you are in a RECHARGE cycle, simply toggling the /CE pins will not reset the /CHG indicator.

Hi Jeff,

I know this is an old post but I am really going through a charger selection process and I'm having difficulties to tell the difference between some of the available IC. I need a dual charger (USB and DC, hopefully a solar charger), able to power path and standalone, thus, no interface with a MCU and able to manage the status outputs with LEDS.

I am right now studying the BQ24165, I will see what to do with IUSB lines, so far I don't think I need this smart signaling.

Let me know if you can give me any advice on this please

Many thanks

bq24165 is the only dual input, standalone switchmode charger we have.  If you need USB compliance than you might need to add a bq24392 to help determine where to set the IUSBx pins.  We also have a dual input linear charger, bq2403x family.

If one input is solar, you can use the VINDPM feature to set that input's regulation point to the solar panel's maximum power point.

Does your battery have a thermistor?  If so, then I suggest change to bq24166 or bq24167.

Jeff,

Thanks for your answers, first of all. 

USB compliance: I won't be having I2C interface so extra IC to control IUSB are kind of discarded. I think it could be working on High Impedance Mode (IUSB3=IUSB2=IUSB1=1), limiting of course the spikes and current on the DC IN, in case a solar panel is plugged. Do you see any disadvantage here?

VINDPM feature: I believe (basing on datasheet) I could not use this feature in High Impedance Mode of giving fix values to IUSB's, could I?. Regardless of that, I still don't see why the USB port would crash...

Battery Thermistor: Yes, an independent thermistor will be mounted, as near as possible to the battery. Got it, the BQ24165 lacks the TS pin for the thermistor.

Thanks again for your help

When in HI-Z mode the charger's buck converter and other internal circuirty is turned off, the battery FET between SYS and BAT is turned on and any load on SYS is powered from the battery.

Even without I2C interface, you can perform primary and secondary detection on the USB rail using the D+/D1 lines and determine the USB current capability to be USB BC 1.2 compliant.  Ideally you would  then enumerate from a USB host and then change the current using GPIO pins to control the IUSBx pins. 

Regardless, you can use the IUSBx pins to set an expected input current maximum from your source attached to the USB input, maybe 100mA or 500mA.  The VINDPM monitors the IN pin and further reduces the input current limit from maximum set by the ILIM pin resistor in the event the IN voltage starts to collapse because too much current is being pulled from it, which will happen for a high impedance output source like a solar panel.

I see. So, in HI-Z mode with CE enabled, it would be charging battery and powering from battery? I may not be understanding the logic into that... 

The MCU is really short on GPIO and besides I2C, we have not many GPIO to use. That is why I am trying to configured the IC somehow fix, but flexible enough to handle USB and DC IN in the best possible manner. If you see a better configuration of the BQ24166, do not hesitate suggest it.

Appreciate your help

In HiZ mode, charging is disabled regardless of CE pin.  The load is powered from the battery.

You can configure USB input current limit using the IUSBx pins and use the VINDPM feature from allowing the USB to drop below the internally fixed VINDPM levels.  This is not 100% USB compliant but should operate okay.

Jeff, not sure if I've got that:

I thought that setting IUSB3-1=001, it would cause the input current drop to 500mA whenever the input voltage drops below 4.44V. I don't see how to lower that threshold while maintaining the IUSB pin fix.

Other thing is: do I have to use external FET to power load from battery (when no IN detected) or the BQ shorts SYS and BAT (I believe with a higher impedance path) itself?

Last question (true!): given that the load will rarely drain more than 60mA and the solar conditions vary, I think that setting Vin-DPM close to Vuvlo (3.8-4V) would allow charging a battery the maximum percentage of time. Am I wrong?

Many thanks for you help  

Setting IUSB3-1=001 sets the input current limit to a maximum of 500mA regardless of how much the input source can supply as long as V(IN) >>4.44V.  If the input voltage starts to drop to 4.44V, the IC will automatically start reduce its input current limit below 500mA so that the input voltage does not fall below 4.44V.

In HIZ, the internal BATFET is shorted, tying SYS=BAT.  If you want an even lower resistive path from BAT to SYS you can add an external FET, driven by BGATE, which will be in parallel with the internal BATFET but only during HiZ mode.

You typically set the VINDPM to 75% of the panel's open circuit voltage to get maximum power. 

The IC also has sleep mode which turns off the charger if the voltage at IN is within 100mV of the battery voltage.  Therefore if you want to fully charge a 4.2V lithium ion battery, your solar panel's open circuit voltage x 75% needs to be >=4.2V.  The minimum voltage to which we have characterized the VINDPM function is 4.2V.