BQ25890: Why VSYS is only 1.87V with a 4.7Om load?

Frank,

Please try to power up the charger first without load. Connect the load after the system voltage reaches 3.65V. The charger soft start feature limits the input current to 200mA before the Vsys reaches 2.2V. More details can be found on the BQ25890 datasheet (version SLUSC86C) page 21.

Regards,

Eric

Hi Eric,

Thank you very much for reply.
Case1: VBUS is connected and Battery is connected

Case2: VBUS is off and Battery is connected

Case3: VBUS is connected and Battery is off

Question 1: "The charger soft start feature limits the input current to 200mA before the Vsys reaches 2.2V" If that, the Vsys can not drive the 4.7Om load in all the three cases. But in fact, only in Case3, the Vsys gets 1.87V. And in the other 2 Cases, the Vsys gets about 3.6V and drive the load normally.

        So, I think the Page 21 is not related with the problem. Am I right?

More description: In my application with my board, when WIFI disconnected, the board takes 250-300mA current with 5V supply. And when WIFI connected, the current consume increase to 350-400mA with 5V supply.

         In Case 3, the application works well when WIFI disconnected. But when connecting to WIFI, the DSP is reset, and with the oscilloscope I find the Vsys has a drop.

         In the above figure, the blue channel is Vsys and the yellow channel is 1.8V for DSP. When the board try to connect to WIFI, the current increase and the bq25890 cannot afford sufficient current, so the Vsys drops.

         In Case 1 and Case 2, the application works well both when WIFI disconnected and WIFI connected.

Question 2: According to Figure 5 and Figure 6 in the datasheet. I think the 2 figures tell us that the Vsys can provide up to 3A current in Case 1 and Case 2. Am I right?

Question 3: But the datasheet doesn't tell us the current limit in Case 3 or give any figure to describe that. What is the maximum current the Vsys can drive in Case 3?

     Thank you so much.

     B.R.

Frank

Hi Jeff F,

    “More description: In my application with my board, when WIFI disconnected, the board takes 250-300mA current with 5V supply. And when WIFI connected, the current consume increase to 350-400mA with 5V supply.”

    The board is charged with a SDP USB port and the current is measured using a USB current meter. And the ILIM pin resist is 220Om, which means the ILIM pin current is 1.8A. The “ LOWER of setting per D+/D- pin OR the ILIM pin setting” is 500mA.

    In case 3, when my board is powered up, the WIFI is not connected by default, and the current is 250-300mA (>200mA), and the SYS voltage is 3.65V-3.7V. So I think it does not matter with the startup things described in page 21 in the datasheet. And when the software connet the WIFI, the current increased and the USB current meter showed the current increase to about 340mA and then drop back to 220mA. The current drop matches the SYS voltage drop in time.

   And in case 1 and case 2, when WIFI is connected, everything works normally.

   So, I think the SYS voltage drop does not matters with the ILIM. I think maybe in case 3, the bq25890 sys can only provide about 300mA current. But I didn't find any information about the maximum current in case 3. Can you help to confirm that and give the maximum current in case 3?

   The figure 5 and figure 6 shows the Current VS Voltage curve for case 1 and case 2. Not case 3.

   Thank you so much.

   B.R.

Frank

Frank,

It is a buck converter. If VBUS=5V and IINPDM=500mA then using efficiency power balance (eff=Po/Pin), SYS should be able to supply ISYS=5V*0.5A*0.90/3.65V =616mA. What is the voltage at charger VBUS pin? Is dropping due to high resistance connection to USB port or due to meter?

Hi Jeff,

When WIFI is not connected, the current is 250-290mA, the VBUS voltage at VBUS pin is 4.97V.

When connecting WIFI, VSYS drops, but the VBUS pin never drop.

For test, when connect the VSYS with a 4.7 Om resist load, the VSYS is 1.65V, but the VBUS pin is 5.03V.

Frank

Frank,


Can you send the schematic for the bq25890, including inductor part number? Do you have a way to read the I2C registers when this is happening, including an ADC measurement? The status registers will tells us what is happening. If not, then can you send oscope plots of VBUS, SW and SYS when this occurs?

Hi Jeff,

     The schematic is below:

   The inductor I use is SPM3020T-2R2M-LR 

   And althrough the I2C is connected to DSP, but the DSP is reset when the SYS drops. I cannot get the register when this happens.

   This is the VSYS (blue) vs the SW (yellow) when this happens:

    This is the VSYS (blue) vs the VBUS (yellow) when this happens:

 

       Thank you so much!

Frank

Frank,

Can you zoom in on the SW pin when SYS drops out to show the squarish waveform like shown in d/s figure 54? Do you have a way to measure the charger's input current? I strongly suspect the charger is input current limited.

Hi Jeff,

     Thank you very much.

     The 2 figures below shows the SW (yellow) and SYS (blue).

     Figure 1, the WIFI is not connected. Figure2, the SYS drops when connecting to WIFI (I set the trigger level to 2.2V, falling edge). Since the SYS drops in several ms, you can see the SYS is a straight line in 100ns scope.

Figure1:                     

Figure2:

The auto measurement in Figure1 is not acurate because of the over shot. The cursor measurement is:

                VSYS                     Pos Width            Neg Width

Figure1    3.7V                       528ns                   160ns

Figure2    2.2V                       352ns                   336ns

I use another 2A charger and use the USB current meter to monitor the current. When the battery is on, the current is 1.08A. So I don't think the problem is because of the input current limited.

   B.R.

Frank

 

Frank,

Can you remove the USB meter and short D+=D- and retry?

Hi Jeff,

     Thank you.

      I ordered a BQ25890EVM-664 from TI and today I got it.

      TEST1. I use a 4.7Om load for test.

      1. When no load is put at the VSYS, the VSYS is 4.2V. (D+ D- not shorted)

       2. Then, I put the load onto the SYS while keeping the board powered by the VSOURCE. The VSYS drops to 1.82V which is similar to the phenonmen that happened on my board (1.75V). Here now, the VSOURCE keeps on 5.08V and the current is 0.18A.

 

   3. I also repeat the above experiment with D+ D- shorted. The exactly same phenonmen happened, the VSYS droops to 1.82V.

  4. You can see that the Vsource did not drop. And the Vsource is a 2A charger and as I said, when the battery is put on, the meter shows the current is 1.08A.

      And I also tried the experiment without the meter. The exact phenonmen happened.

5. If a 10Om load is used. The VSYS is 4.19V and everything is OK.

    So, I think the bq25890 VSYS can not provide > 4.2/4.7=893mA current when battery is off.

    Question1. What is the maximum current that the VSYS can provide when the battery is off?

    TEST2.  Compare my board with BQ25890EVM-644.

    Both the boards can not drive a 4.7Om load.

    When the load is 10Om, the VSYS on my board is 3.64V while the VSYS on EVM is 4.19V.

    Question2 : What is the difference between my board and EVM?

    Question3 : The STATE LED on my board blink once at a time, while the STATE LED on the EVM blinks much quickly. Why?

 

     B.R.

Frank

Frank,

I have used the bq25890 to provide well above 2A from 5V/3A input so I know it is capable. To summarize:
1. At startup with SYS < ~2.0V, input current is limited to 200mA so the charger will not startup into 4.7ohm.
2. After startup, the input current limit is clamped to the lower of the setting per the ILIM resistor or setting in the IINDPM register. With the 890, the IINDPM register is configured by D+/D- pins. With D+/D- shorted, IINDPM is 3A but, unless you disable ILIM resistor via I2C or also change resistor to set it to 3A, the ILIM resistor clamps the input current below 3A. If I recall correctly, the EVM ILIM resistor has a potentiometer in parallel with fixed resistor. If you increase the potentiometer to the maximum, so that help?
3. If the load transient caused by quickly added such a large load causes the SYS voltage to drop <~ 2.0V, the charger could get confused and think it is in startup, which returns you to condition one above. In that case, you will need to add additional capacitor from SYS to GND that can absorb more of the transient and prevent such a droop.

With the EVM, you can use the bqStudio I2C software to read the charger's status register, which might tell you the problem. Do you have an EV2300 or EV2400 I2C box to use with the software?

Hi Jeff,

    Thank you very much.

    I find that the JP10 is shorted with a jumper on the EVM, and the VSYS is 4.2V. Remove the jumper in JP10, the VSYS become 3.65V.

    In my test, the RLIM is 130Om on the EVM (the R19 potentiometer is set to 0Om). So the ILIM is 355/130=2.73A.

    In my test, the 4.7Om load is in parrel with a 1uF cap. With the 4.7Om load, the VSYS is 3.13V.

    And with the 10Om load, the VSYS is 3.65V, when I open the power first, and then connect the load to VSYS manually.

    

With the 4.7 Om load:

The adapter is recognized as a SDP, and the current limit is 500mA. 3.65/4.7*3.65/0.9/5=0.63A.0.5A. 

That's why the VSYS drops. And if I set the current limit to 3.25A, the VSYS can be 3.65V even with the 4.7Om load.

Thank you for your advice.

But I still have some problem with the EVM:

Problem1: With a 2A adapter and D+/D- shorted on the EVM board, the adapter is still recognized as a SDP. (current meter is not used). But on my board, I did the experiment many days ago and it's recognized as DCP.

Problem2: With the battery on, the adapter never charges the battery. But with my board, the adapter charges the battery with 1.08A.

For the above 2 problems, I've also tried other adapters, but always SDP and never charge.

I have not found the reason from the registers, could you help. Thank you so much.

B.R.

Frank

Frank,

First, please check the EVM to ensure that it has a bq25890 installed and not a bq25892.  Then, confirm that JP1 shunt is not installed and JP2 shunt is connected between pins 1 and 2.  If you are using a power supply, you can install the shunt on JP5 to short D+=D- and simulate a DCP.  If you are plugging in a DCP adapter, use J5.  If that doesn't work, try flipping the S2 switch.

Hi Jeff,

        I use a power supply, set 5V output and 2A current limit. The S2 is switched to the right position which makes D+ D- shorted. The battery and a 10Om load is connected to the VBAT and VSYS receptively. The Rlimit is adjusted to the smallest value 130Om, so the Ilimit set by the Rlimit is 335/130=2.577A.

The connection is shown below:

The Vsource is recognized as a DCP, and the regisiters are shown below:

If the power supply is off, the VSYS is 3.7V which means the power now is supplied by the battery (VBATT is about 3.7V for this half full battery).

Question 1: Why the Input Current limit is only 500mA?

Question 2: Why the state is still not charging? The Isource shown in the power supply is 320mA, but if the battery is removed, the Isource is 297mA. So the power supply drives 20mA into the battery. I wait for 10 minutes, the Isoure is still 320mA and the state is still not charging.

 

Then I change another battery and switch the power supply off and on. This battery is nearly full, the VBATT is about 4.124V as shown below,

The state is still not charging, and the input current limit is still 500mA.

But this time, the Isource is 370mA, which means about 70mA is driven into the battery. But the state is still not charging.

Please help. Thank you.

B.R.

Frank

Hi Ning T,

       Thank you very much. Now I have no problem with the EVM board.

       But with my own board, there is still a strange phenonmen.

       I use a 5V2A power supply and with D+D- shorted.

       When the battery is on, the bq25890 can recognize the source as DCP when power up.

       But when the battery is not connect, the bq25890 recognize the source as "Unknown adapter (500mA)".

       This is the registers in the two conditions with my board. I did the same experiments with the EVM, it can recognize the source correctly in both condition.

       Can you help me to find the problem with my board?

       Thank you.

Condition	No Battery		With Battery
ISource	268mA		557mA
REG00	0x48	IINLIM=500mA	0x7F	IINLIM=3.25A
REG01	0x6		0x6
REG02	0x5D		0x5D
REG03	0x1A		0x1A
REG04	0x20		0x20
REG05	0x13		0x13
REG06	0x5E		0x5E
REG07	0x9D		0x9D
REG08	0x3		0x3
REG09	0x44		0x44
REG0A	0x73		0x73
REG0B	0xA7	101:Unknownadpter （500mA） NotCharging In VSYSmin regulation	0x76	011:DCP(3.25A)   FastCharging Not In VSYSmin regulation
REG0C	0x5	TS Cold	0x0
REG0D	0x12		0x12
REG0E	0x0	VBAT=0V	0x4D	VBAT=3.844V
REG0F	0x44	VSYS=3.664V	0x4D	VSYS=3.844V
REG10	0x7F		0x49
REG11	0x96	VBUS=4.8V	0x93	VBUS=4.5V
REG12	0x0	IBAT=0mA	0x5	IBAT=250mA
REG13	0x8	Not in IINDPM IDPM_LIM=500mA	0x4A	IINDPM IDPM_LIM=600mA
REG14	0x1D	Optimization is in progress	0x5D	Maximum Input Current Detected

And the following table shows the difference between my board and EVM board, when battery is not on.‘

It's strange that, although the battery is not on, the EVM still get VBAT=4.184V and VSYS=4.224V. And I checked the voltage with  multimeter, that is true. 

And the STAT LED blinks very very quickly on the EVM, about 5Hz, not 1Hz. Why?

I think when the battery is removed, the VBAT should cut off, and the VSYS would be 3.65V according to the datasheet.

Condition	My board		EVM board
REG00	0x48	IINLIM=500mA	0x7F	IINLIM=3.25A
REG01	0x6		0x6
REG02	0x5D		0x1D
REG03	0x1A		0x1A
REG04	0x20		0x20
REG05	0x13		0x13
REG06	0x5E		0x5E
REG07	0x9D		0x9D
REG08	0x3		0x3
REG09	0x44		0x44
REG0A	0x73		0x73
REG0B	0xA7	101:Unknownadpter （500mA） NotCharging In VSYSmin regulation	0x7E	011:DCP （3.25A） Charging TerminationDone Not in VSYSmin regulation
REG0C	0x5	TS Cold	0x80
REG0D	0x12		0x12
REG0E	0x0	VBAT=0V	0x5E	VBAT=4.184V
REG0F	0x44	VSYS=3.664V	0x60	VSYS=4224V
REG10	0x7F		0x51
REG11	0x96	VBUS=4.8V	0x98	VBUS=
REG12	0x0	IBAT=0mA	0x0	IBAT=0mA
REG13	0x8	In VINDPM Not in IINDPM IDPM_LIM=500mA	0x3F	Not in VINDPM Not in IINDPM IDPM_LIM=3.25A
REG14	0x1D	Optimization is in progress	0x1D	Optimization is in progress

I think the only difference between my board and EVM is that I use a bq27441 to monitor the volume of the battery.

I removed the bq27441 from my board, but it doesn't change anthing.

Hi Ning T,

     Thank you very much. Following your advice, the VBAT will not recharge when battery is removed.

     As shown by the last 2 posts, my board has a strange phenomenon which is different from the EVM.

     My board can recognize the DCP correctly when the Battery is on. But cannot recognize the DCP when the Battery is off. Could you give me some advice on what may cause that?

     Thank you so much!

     Best regards.

Frank

For your board without battery, REG0C=05 (TS cold). Do you know why? Thanks!

Hi Ning T,

     Thank you very much for reply.

     Because there is a JP10 on the EVM board, while my board do not have. When the 10k is connected, TS will be normal.

     When I remove the JP10 jumper of the EVM, the EVM is also TS cold. But the EVM can still recognize the DCP correctly.

     B.R.

Frank

Hi Ning T,

       I've found the cause of my problem.

       In my schematic, I use a TS3USB221 to switch between the Host D+D- and the bq25890 D+D-, which is similar to the typical application example shown in bq25890's datasheet.

       But the TS3USB221 is powered by VCC_SYS. So when the battery is off. When the USB is plugged in, the VCC_SYS is 0, so the D+D- does not get through. So the bq25890 cannot recognize the USB connectly.

       So, I should use a USB SWITCH whose VCC can be 5V to 14V, since there are many 9V fast USB charger. But I find there is no such USB SWITCH which can tolerate 9V or 14V.

       So, the solution is that I can use a 5-14V input 3.3V output LDO to power the TS3USB221. I think hat will make bq25890 recognize the USB source correctly when the battery is off. Do you think the solution is OK, could you give me some advice?

       Thank you.

    B.R.

Frank

Hi Ning T,

     Thank you so much for your help!

     Best regards!

Frank