BQ25756E: Schematic Review Request - 140W USB-PD 3.1 Charger for 6S Li-ion

Hello Merri,

Have you reviewed your schematic with the schematic and layout checklist, yet?

We also have an app note to help MOSFET selection, a reference design for 100W USB-PD charging, and 240W USB-PD charging.

Merri Jones said:

Does the BQ25756 handle thermal throttling autonomously via the TS pin, or does this require MCU intervention?

This is automatic. The BQ25756 will follow the JEITA profile when the thermistor gets hot.

Merri Jones said:

Is this "Hybrid" approach safe? Will the BQ25756 stability/compensation loop handle the transition from 5V (low power) to 28V (high power) with the 3.3microH inductor selected for the higher voltage range?

Yes, this is safe. Make sure to select the right switching frequency for the inductor. What is the switching frequency for your system?

Merri Jones said:

The calculator flags 3.3microH as close to the minimum. Given the wide input range (5V–28V) and wide battery range (16V–25.2V), is 3.3microH the optimal trade-off for a 140W target, or should I move to 4.7microH to reduce ripple at the expense of transient response?

The 3.3µH and 4.7µH will both work good for the transient response. The main tradeoff between the 2 inductors will be the efficiency. Does your application have a size limit for the inductor.

Let me know if you have reviewed the schematic with the checklist. Once you have gone through the checklist, I can do a final schematic review on my side.

Best Regards,
Ethan Galloway

Hi Ethan,

Thank you for the guidance. I have reviewed the Schematic and Layout Checklist against my design and updated the schematic accordingly. I am holding off on starting the PCB Layout until you give the schematic a final green light.

Design Updates:

Inductor & Frequency: Moved to 4.7µH and lowered switching frequency to 400kHz (R_FSW = 66.67kΩ (should be correct)) to prioritize thermal efficiency in our enclosed form factor.

Thermistor: Configured the JEITA network (RT1 = 5.23kΩ, RT2 = 30.1kΩ) for a standard 103AT NTC.

Two Clarifying Questions:

The CE (Charge Enable) Pin: The checklist implies this should be driven or pulled High, but since CE is Active Low, I have placed a 10kΩ Pull-Down resistor to GND.

Reasoning: I want the default state to be "Charging Enabled" so the device can recover from a dead battery even if the MCU has no power to drive the pin.

Question: Is this safe/recommended, or should I add a hardware delay?

MCU Connections: The current schematic page focuses on the BQ25756 power stage. Do you need to see the MCU page to review how I am handling the I2C lines (SDA/SCL) and status signals (STAT/PG)? I want to ensure the logic levels and pull-ups match your expectations.

Also since I am fairly new to power hardware can you explain the reasoning for the multiple decoupling capacitors in the reference designs (which I copied). 

Finally will your review also cover the USB controller and the rest of the implementation?

I have attached the updated Schematic (Rev 1.2) for your review.Rev1_2.pdf

Best

MJ

Hello Merri,

Merri Jones said:

Reasoning: I want the default state to be "Charging Enabled" so the device can recover from a dead battery even if the MCU has no power to drive the pin.

Question: Is this safe/recommended, or should I add a hardware delay?

Yes, this is safe.

Merri Jones said:

MCU Connections: The current schematic page focuses on the BQ25756 power stage. Do you need to see the MCU page to review how I am handling the I2C lines (SDA/SCL) and status signals (STAT/PG)? I want to ensure the logic levels and pull-ups match your expectations.

These connections look good.

Merri Jones said:

Also since I am fairly new to power hardware can you explain the reasoning for the multiple decoupling capacitors in the reference designs (which I copied). 

There are multiple reasons to have multiple different decoupling capacitors;

• Multiple capacitors can have a lower ESR in parallel
• Multiple capacitors are more resilient if one capacitor fails.
• Multiple capacitors are shorter and can be easier to fit on a board compared to one big capacitor

You might have already seen this, but we have an app note for the capacitors.

Merri Jones said:

Finally will your review also cover the USB controller and the rest of the implementation?

I've looped in the USB controller team to review this portion.

Your schematic looks good. I have 3 thoughts:

• It looks like you have about 260uF on the input and the output. This will work for the BQ25756E, but you might be able to save money and board space by only using 160uF.
  
  
• Make sure the VAC pins are shorted together close to the BQ25756.
  
  
• I would also recommend adding a DNP 0 Ohm resistor to short DRV_SUP to REGN. This resistor will help with debugging.

Best Regards,
Ethan Galloway

Thank you Ethan.

The app note on capacitors is not showing up as existing when I click that link. I will add a 0 ohm resistor to short DRV_SUP to REGN. Are there any places that you see where I can save costs without compromising functionality? This is crucial or even removing parts as a whole would be great.

Ideally when plugged in I would like the mcu to be able to power up so how would we support this, would it be (2) that you mentioned?

Additionally I read that pin for pin the TPS25751 is exactly identical drop in replacement for the TPS26750 so should I just use that instead? Appreciate all the feedback and guidance thus far.

Hello,

Merri Jones said:

Ideally when plugged in I would like the mcu to be able to power up so how would we support this, would it be (2) that you mentioned?

I would recommend connecting the TPD device to LDO_3V3 to ensure that the PD can negotiate to the desired sink voltage to charge/power the system when the battery is dead. 

Merri Jones said:

Additionally I read that pin for pin the TPS25751 is exactly identical drop in replacement for the TPS26750 so should I just use that instead? Appreciate all the feedback and guidance thus far.

The TPS25751 is the same package but the TPS25751 is intended for SPR (20V) while the TPS26750 is intended for EPR (48V).  If you use the TPS25751 because the EPR is not needed then the TPS4S480 is also not needed.  You can look to the TPS25751EVM as an example which replaces the TPD4S480 with the TPD4S201.

https://www.ti.com/lit/pdf/SLVUCP9

Regards,
Chris