BQ25792: ADC IBAT and IBUS channels are intermittent during OTG operation.

Jim McGarvey

Part Number: BQ25792

IBAT and/or IBUS periodically report incorrect values during OTG operation. The actual battery and VBUS currents are steady (measured outside the part) as are all the other reported ADC channels, and I don't see any faults or changes in other registers. The incorrect currents are either very low values, or half the normal value and opposite polarity, like this...

Normal: -1800 to -1900

Incorrect: -10 to +10, or +900 to +050

Below are register dumps in normal and false low current times, and also an i2c trace that looks fine in both cases.

VSYSMIN   00: 7000     VINDPM    05: 3600     PCHG      08:01
VREG      01: 8400     IINDPM    06: 3300     TERM      09:05
ICHG      03: 4200     VOTG      0B: 12000     RCHG      0A:63
                        IOTG      0D:2D         TIMR      0E:3D
ChgCtl 0 0F:82         ChgSts 0 1B:87         TEMP      16:C0
       1 10:00                1 1C:0E         NTC0      17:7A
       2 11:00                2 1D:01         NTC1      18:01
       3 12:40                3 1E:00         ICO       19:00
       4 13:29                4 1F:00
       5 14:04
FltSts 0 20:00         FltFlg 0 26:00         FltMsk 0 2C:00
       1 21:00                1 27:00                1 2D:00
ChgFlg 0 22:00         ChgMsk 0 28:00
       1 23:00                1 29:00         AdcCtl    2E:B8
       2 24:00                2 2A:00         AdcDis    2F:00
       3 25:00                3 2B:00         AdcDis    30:00
IBUS      31: -1807     IBAT      33: -1955     TS        3F:     0
VBUS      35: 7589     VBAT      3B: 7772     TDIE      41:    57
VAC1      37: 7559     VSYS      3D: 7810     D+        43:     0
VAC2      39: 7543                             D-        45:     0
DPDM      47:00         Part      48:08

VSYSMIN   00: 7000     VINDPM    05: 3600     PCHG      08:01
VREG      01: 8400     IINDPM    06: 3300     TERM      09:05
ICHG      03: 4200     VOTG      0B: 12000     RCHG      0A:63
                        IOTG      0D:2D         TIMR      0E:3D
ChgCtl 0 0F:82         ChgSts 0 1B:87         TEMP      16:C0
       1 10:00                1 1C:0E         NTC0      17:7A
       2 11:00                2 1D:01         NTC1      18:01
       3 12:40                3 1E:00         ICO       19:00
       4 13:29                4 1F:00
       5 14:04
FltSts 0 20:00         FltFlg 0 26:00         FltMsk 0 2C:00
       1 21:00                1 27:00                1 2D:00
ChgFlg 0 22:00         ChgMsk 0 28:00
       1 23:00                1 29:00         AdcCtl    2E:B8
       2 24:00                2 2A:00         AdcDis    2F:00
       3 25:00                3 2B:00         AdcDis    30:00
IBUS      31:    -6     IBAT      33:     0     TS        3F:     0
VBUS      35: 7599     VBAT      3B: 7757     TDIE      41:    57
VAC1      37: 7550     VSYS      3D: 7841     D+        43:     0
VAC2      39: 7545                             D-        45:     0
DPDM      47:00         Part      48:08

over 4 years ago

0 Jeff F over 4 years ago

TI__Guru**** 165886 points

Hi Jim,

That is unexpected. Does that IBUS read correctly during forward/charge mode? Is the OTG current limit feature working as expected? Is this on the EVM or your board? If your board, are the PMID and VBUS 0.1uF decoupling capacitors placed as close to IC pins and GND as possible with PMID cap not via'd?

Regards,

Jeff

0 Jim McGarvey over 4 years ago in reply to Jeff F

Intellectual 490 points

IBUS and IBAT work correctly when charging.

OTG current limit work correctly. In the case above, there is a 4 ohm load on VBUS and the part drops the voltage from the VOTG of 12V to keep the current under 2A as programmed.

This is on my board. The PMID and SYS caps are placed like the EVM. I did not think the VBUS cap is as critical since that is not a switching node (and in my test, the load was a resistor so no noise on VBUS).

0 Jeff F over 4 years ago in reply to Jim McGarvey

TI__Guru**** 165886 points

Hi Jim,

I like your new layout, even better than the one I did for the EVM!

To my knowledge, no other customer has mentioned this issue. What were your settings for ADC rate, average, resolution and average start? If you change any of those, does that correct the issue or make it worse?

Regards,

Jeff

0 Jim McGarvey over 4 years ago in reply to Jeff F

Intellectual 490 points

You can see the ADC settings in the register dump above. I will try changing them shortly and let you know.

Thanks for the compliment. The layout works really well thermally, which matters with two chargers on a 75 x 105mm board in a small box, charging at 4A each!

0 Jim McGarvey over 4 years ago in reply to Jeff F

Intellectual 490 points

IBAT and IBUS work correctly when averaging is disabled.

ADC resolution setting has no effect.

Changing average start has no effect, and the bit always reads zero. Not sure if that's expected.

Disabling all channels except IBUS and/or IBAT has no effect.

0 Jeff F over 4 years ago in reply to Jim McGarvey

TI__Guru**** 165886 points

Hi Jim,

Thanks for the expert debug. I will test a few units on the bench. To my knowledge, we have not seen this issue before. Does it happen on all units or only a few?

Regards,

Jeff

0 Jim McGarvey over 4 years ago in reply to Jeff F

Intellectual 490 points

I've only tried two units so far. Both act the same.

0 Jim McGarvey over 4 years ago in reply to Jeff F

Intellectual 490 points

Jeff, I also tried turning down the OTG current as low as 200mA and that did not help.

+1 Jeff F over 4 years ago in reply to Jim McGarvey

TI__Guru**** 165886 points

Hi Jim,

Thanks for the update! I have tested a unit on the bench and it shows the same issue. I am following up with our design and validation teams but it seems that we will need to make a change to the datasheet in the future that does not allow running average when measuring current in OTG mode. Hopefully we can fix this issue for future devices using this ADC.

Regards,

Jeff

0 Jim McGarvey over 4 years ago in reply to Jeff F

Intellectual 490 points

Thanks, Jeff. I appreciate your honesty! This is something I can workaround with no problem. I only wanted to investigate in case it was a problem with my design.

Power management

Power management forum

BQ25792: ADC IBAT and IBUS channels are intermittent during OTG operation.