BQ25892: Can't seem to get the charger working

Hi Greg,

The datasheet schematic on page 52 shows the minimum components required for operation. PMID and ILIM cannot be left floating. SYS needs at least 20uF and BAT needs at least 10uF of capacitance. VBUS needs at least 1uF. With ILIM floating, the charger's input current is set almost to zero.

Thanks Jeff!

I mistook some of the documentation "The ILIM pin function can be disabled when EN_ILIM bit is 0." as being able to leave it unconnected if I disable this function. I will hook up this pin and the others as you suggest and report back.

For the larger capacitance needs on BAT and PMID, can I use polarized electrolytics? Second, any issues with a non-polarized electrolytic 4.7uF for C1 on REGN? (I made an error in my schematic, I will also update and post a revised one this evening).

Hey Jeff!

Here is the latest schematic. I've fixed a few errors and added additional capacitance as you suggested.

Unfortunately, I'm still not able to get a charge current. The cell is 3.7V. I've set REG0 to 00100000, which is disabling the ILIM pin, and setting charge current to 1600mA.

Status
------
VBUS Status : Adapter (3.25A)
Charging Status : Charging
Power Status : Good
VSYS Status : Not in regulation
Thermal Status : Not in regulation

Battery voltage is 3524mV
Charge current is 0mA

Greg,

Are you able to communicate via I2C? If so, please perform a read and send the register contents. If not, what does an oscilloscope plot of the SW node show? What is the current rating of the 1uH inductor?

Hi Jeff,

Yes, I2C works. registers below.

Inductor is Wurth 7447462010 - rated at 8.5A.

Scope plot of SW also attached below.

Thanks for your continued assistance :)

Register  0 has value 00000000
Register  1 has value 00000110
Register  2 has value 01001100
Register  3 has value 00010000
Register  4 has value 00001000
Register  5 has value 00100011
Register  6 has value 01011010
Register  7 has value 11101101
Register  8 has value 00000011
Register  9 has value 01000000
Register  a has value 01110011
Register  b has value 01010110
Register  c has value 00000000
Register  d has value 00001110
Register  e has value 00111001
Register  f has value 00111010
Register 10 has value 00111100
Register 11 has value 10010101
Register 12 has value 00000000
Register 13 has value 00000000
Register 14 has value 00000101

Greg,

If you have a current meter in series with the battery, I recommend removing it. If you can't do that, then disable autoranging. Your registers report 100mA input current limit and USB CDP adapter, which is odd. There is no automatic setting for 100mA. If you increase the input current to 600mA or higher, does your circuit function correctly? Regarding the SW plot, can you trigger on the cleaner edges at the top or bottom of the waveform?

This has been with the current meter removed.

Attached another SW plot, not sure if it's any better.

RE:100mA, I think you were looking at the first values for the BQ25890. I am using 892, which has

bq25892
000: No Input
001: USB Host SDP
010: Adapter (3.25A)
111: OTG
Note: Software current limit is reported in IINLIM register

010 = 3.25A. However, IINLIM doesn't look so hot. REG00 is 00000000b. With an offset of 100mA, you are correct it's being current limited, assuming it respects those values.

Even if I program REG00 to 800mA, nothing seems to happen. Subsequent reads of REG00 confirm that I set it 800mA without any effect.

Greg,

Your waveforms do not look correct. The bottom should be flat go to ground and top should also be flat and reach VBUS. Are you certain that the ground is properly soldered. It might be easier to evaluate using a bq25895EVM.

Is SYS shorted to ground? There is short circuit protection that limits the current.

Hey Jeff,

Yeah after spending a few more hours last night I came to the conclusion I don't think it's a configuration/software issue but a hardware issue.

The EVM is a bit out of my price range for this project. I'll try to do some hardware debugging to clean up the circuit and perhaps just go through a few rounds of prototyping. I can also lean heavily on the EVM reference design, and put in a few options for simplifying the circuit at assembly time.