BQ25895: BATFET turns off abruptly at battery temperature of 45DegC.

Hi Ashish,

We will take a look and provide an update.

Regards,

James

Thank you for looking. When the charging is suspended at the published VHTF threshold on datasheet page 9, the battery is also being isolated in the power path and this is cutting off power to the system if there is no external DC source to provide power.

Hi Ashish,

After testing on the EVM, I was still able to draw current from the battery after the TS threshold was reached. However, without a valid VBUS voltage VSYS will fall slightly below the battery voltage. If OTG mode is on, PMID should be raised to the OTG regulation voltage automatically if all the conditions in Section 8.2.5 of the datasheet are met.

How low does VSYS drop once the temperature threshold is reached?

What is your ISYS (current draw on SYS node)

Can you provide a scope capture showing VBAT (BAT pin voltage close to IC), IBAT (battery current), VSYS, and VBUS?

Regards,

James

If there is no external power adapter connected to the device, the battery continues to provide power after the VHTF threshold is reached at high temp.

However, if the battery is charging and then the threshold is reached, the charging is suspended and device blacks out as soon as external power adapter is disconnected.   Have you tried this use case on the EVM?

I will provide you the sope capture you recommended. I have a device running in the temp chamber monitoring VSYS, VBAT , VREG and TS pins, so I need to solder some more lines to the test points  to capture the parameters you are looking for.

Hi Ashish,

I will grab some scope captures for you by tomorrow so you can compare with your findings.

Regards,

James

Hi Ashish,

Due to worsening winter weather, I will not be able to get into the lab today. I will update you as soon as I can.

Regards,

James

Hi James,  Once temp threshold is reached, the VSYS drops to 2V. The steady state Isys current is 250mA, could peak upto 500mA for short periods. I ve attached a scope capture showing Yellow=VBAT, Pink =IBAT, Green=VSYS, Blue=VBUS. This is at room temperature. At trigger point, external 5V adapter is removed and VBUS starts falling. Will get you a capture with same parameters at high temperature charge suspend threshold.  

Hi Ashish,

I will be able to take a look today and provide an update.

Regards,

James

 Same signal capture as above, except at high ambient temperature of 43C. It shows that the VSYS cuts off a few milliseconds after VBUS is removed.

Hi Ashish,

Here is a scope capture I took with the TS pin held outside the temperature threshold. At the trigger point, I am disconnecting the adapter to see whether VSYS crashes at the same time as VBUS. 

At the beginning of the capture, IBAT is 0A because of the TS pin threshold, but once the adapter input is removed, the BATFET opens automatically so that the battery can supply the SYS load.

VSYS will drop to just below the battery voltage since the battery is discharging through the BATFET.

Throughout this sequence, I am keeping the TS pin outside the normal temperature range. However, the BATFET does not prevent the battery from supplying SYS once the adapter is removed.

ISYS = 250mA at all times and VBAT = 3.27V.

Can you provide all your register values before and after you see VSYS crash?

Regards,

James

Hi James,

Your interesting. How much capacitance do you have on VSYS and PMID pins?  At our end we are still seeing that the BATFET is cutting off the battery preventing from supply Vsys. We confirmed that the issue is not caused by the system by using a 0-10K potentiometer at the thermistor pin input instead of the battery thermistor. We lowered the resistance to the charge suspend threshold , after IC is in charge suspend mode, we removed the adapter and are seeing the same behaviour. The green waveform above is VSYS, Yellow is battery, blue is VBUS and Pink is IBAT.

Is there anything in the register settings that could cause this behavior? Would you like to review our schematic page with the charger IC on it?

We ve also noticed that there is a battery voltage dependency for this condition. The issue does not happen with a low battery, voltage < 3.5V. It happens every time with a higher voltage, we are spending some effort here characterizing the battery voltage threshold at which we start to see the BATFET cutting out. Could you please repeat the same test you performed at your end, but with battery voltage >3.8V and Vbus=5V and an additional 100mA load on PMID (plus the VSYS load) ,  so we have similar test conditions? 

We now have a special version of the code which prints out all register values in the charger IC to a log file, will work on getting the registers before and after across to you today.

Hi Ashish,

Go ahead and share your schematic so I can take a look.

Ashish Shah said:

How much capacitance do you have on VSYS and PMID pins?

I am using the BQ25895EVM to check the behavior. The User's Guide has a schematic located on page 18.

Ashish Shah said:

Is there anything in the register settings that could cause this behavior?

If you have Boost Mode enabled, try disabling the Boost Mode temperature monitoring by setting REG01[7:6] = 11b to make sure this isn't an issue with the transition to OTG mode. If your VBUS is falling below VBAT + VSLEEP, this will trigger Boost mode (OTG mode) assuming all other conditions are met from Section 8.2.5 of the datasheet.

Ashish Shah said:

The issue does not happen with a low battery, voltage < 3.5V

Check to see if the behavior changes depending on if VBAT is above or below BATLOW. This is also related to boost mode behavior.

Ashish Shah said:

We now have a special version of the code which prints out all register values in the charger IC to a log file

Thank you, I'll take a look when those are ready.

Regards,

James

We have set REG01[7:6] to 10b for BHOT2 threshold. This works out to 65C. THe temperature is well below the threshold, so this does not seem to be an issue with transition to OTG mode. Please let me know if the EVM functions without cutting off VSYS when battery voltage is 3.8V or higher. We are interested to know if you were able to duplicate the issue under similar test conditions as us ( more details in my post above)

0246.Charger Schematic.pdf

Hi Ashish,

I took a look at the schematic and I have a couple notes:

• For the inductor L2, our max recommended value is 2.2uH (See the Schematic Checklist here).
• The TS pin should be connected to a resistor divider and NTC thermistor in a typical application. Other configurations should model this behavior as close as possible to match our recommendations. The datasheet assumes a 103AT NTC Thermistor.
  

After testing the EVM under the conditions you mentioned, I saw two behaviors depending on the BHOT / BCOLD bit settings.

1) REG01[7:6] = 10b; VBUS = 5V, VBAT = 4V, ISYS = 250mA, IPMID = 100mA

2) REG01[7:6] = 11b; VBUS = 5V, VBAT = 4V, ISYS = 250mA, IPMID = 100mA

Based on the two different cases, I would suggest checking the REGN and TS pin voltage in a scope capture to see how the percentage of VREGN changes during the overtemperature event.

Regards,

James

Hi James, We are seeing the same behavior that you are observed on the eval board in your above post.We learned over the last few days  that turning OTG on while the charger IC is in charge suspend state ( temperature fault  NTC fault status =110 – TS Hot ) causes BATFET to turn off and Vsys to cut off.  It is the state of turning on the OTG that causes the failure. The failure is not caused by just removing external power unless the charger is configured to turn on OTG automatically when external power is removed. Changing the value of the inductor to 2.2uH does not seem to fix this behaviour.

We also did an experiment where BHOT protection is configured to be active, 65C threshold, temperature of the thermistor pin is simulated to be above the charge suspend threshold of 45C and OTG is manually turned on and off via the OTG_CONFIG bit. We timed turning on the OTG  in two cases, A and B. 

Case A) The charger IC is  sampling the temperature of the battery thermistor. No input on VBUS.

Case B) When charger IC is not sampling temperature of the battery thermistor. No input on VBUS.

The issue of BATFET cutting off is always happening in Case A and never seems to happen in CASE B. 

In the attached scope shot,

               Channel 1 (Yellow)           VBUS

               Channel 2 (Green)            PMID

               Channel 3 (Blue)               VSYS

               Channel 4 (Red)                TS

While just enabling OTG with software while coincident with Temp sampling (and REGN present accordingly), the VSYS cut out. We are not sure why this is happening only when coincident with the sampling. What could be the reason for this behavior ? Curious if you are able to duplicate it on the eval board?

Also, trying to understand why this happens when device is in charge suspend status at thermistor temperature of 45C, even though temperature is well below the BHOT protection setting of 65C. It does not seem to happen when device is NOT  in charge suspend status.

Also, a small typo in my above post, NTC fault status =010 not 110.

The 2nd Scope shot below is with manual enabling of OTG using the external OTG pin on charger chip.While just enabling OTG with hardware while coincident with TS (and REGN accordingly), the BATFET cut out and VSYS dropped. You can see that on the 3rd sampling pulse in the end.

,

Hi Ashish, 

I will be taking this thread over for James for the time being. Thank you for providing the two most recent plots and explaining your finding that the behavior seems to occur when the IC is sampling the TS voltage.

I agree with your statement that the Vsys cutoff behavior appears to occur due to attempting to enter OTG while the IC has a TS Hot Status and Boost Mode thermal protection is also enabled. 

Ashish Shah said:

Also, trying to understand why this happens when device is in charge suspend status at thermistor temperature of 45C, even though temperature is well below the BHOT protection setting of 65C.

To help investigate this can you help to provide all register values from the charger IC before and after the Vsys crash, as well as, a scope capture showing the Vsys, TS, and REGN. This information is needed to help understand what the charger is detecting that causes it to disable boost and the BATFET. 

My hunch is that for one reason or another the device is disabling boost mode due to a BHOT detection. Seeing the register values before and after Vsys shuts down will confirm if this is the case. 

Best Regards,

Garrett 

Hello Garrett,

Thank you for looking into this issue. The TS and REGN lines look  clean on the scope. The VSYS crash occurs when temperature of thermistor rises to 45C and  TS voltage goes below the V(HTF) threshold of 48% i.e. (Voltage on TS pin) / (Voltage on REGN pin) <48%

Register data we captured is as follows-

Disconnecting VBUS:

(000252347) Chg Registers: REG00 0x66, REG01 0x86, REG02 0x11, REG03 0x34, REG04 0x00, REG05 0x11, REG06 0x46, REG07 0x8D

(000252351) Chg Registers: REG08 0x23, REG09 0x44, REG0A 0x03, REG0B 0xC6, REG0C 0x02, REG0D 0x14, REG0E 0x55, REG0F 0x58

(000252352) Chg Registers: REG10 0x30, REG11 0x99, REG12 0x00, REG13 0x26, REG14 0x39

(000252412) BatteryCharger::adjustChargingState - Old Charge State: 0x00 Current: 0x00 Temp: 0x30

(000253521) [DekaLogInfo] Event: 1116 (safetymonitor_task.cpp,line:595)            AC_ABSENT

(000253761) BatteryCharger::adjustChargingState - Old Charge State: 0x00 Current: 0x00 Temp: 0x00

(000254956) Temp (K) 302, AvePow 0, NomAvailCap 4392, FullAvailCap 4478, FullChrgCap 4285, FullChrgCapU 4327, SOH (%) 90, SOHS 1, SOCU 99

(000254958) Status 0x009A, Flags 0x0088, OpConfig 0x25F8, DesignCap 4500, TSVoltagePCNT 0

(000254959) dbg1 5400, dbg2 2980, dbg3 2592, dbg4 2720, dbg5 4241, dbg6 128, dbg7 65494

(000254964) Chg Registers: REG00 0x66, REG01 0x86, REG02 0x11, REG03 0x34, REG04 0x00, REG05 0x11, REG06 0x46, REG07 0x8D

(000254965) Chg Registers: REG08 0x23, REG09 0x64, REG0A 0x03, REG0B 0x02, REG0C 0x00, REG0D 0x14, REG0E 0x00, REG0F 0x00

(000254966) Chg Registers: REG10 0x00, REG11 0x00, REG12 0x00, REG13 0x26, REG14 0x39

Connecting VBUS:

(000263604) Chg Registers: REG00 0x66, REG01 0x86, REG02 0x11, REG03 0x34, REG04 0x00, REG05 0x11, REG06 0x46, REG07 0x8D

(000263608) Chg Registers: REG08 0x23, REG09 0x64, REG0A 0x03, REG0B 0x02, REG0C 0x00, REG0D 0x14, REG0E 0x00, REG0F 0x00

(000263609) Chg Registers: REG10 0x00, REG11 0x00, REG12 0x00, REG13 0x26, REG14 0x39

(000263623) BatteryCharger::adjustChargingState - Old Charge State: 0x00 Current: 0x00 Temp: 0x00

(000264935) Temp (K) 302, AvePow 0, NomAvailCap 4392, FullAvailCap 4478, FullChrgCap 4285, FullChrgCapU 4327, SOH (%) 90, SOHS 1, SOCU 99

(000264936) Status 0x008A, Flags 0x0088, OpConfig 0x25F8, DesignCap 4500, TSVoltagePCNT 48

(000264937) dbg1 5400, dbg2 2980, dbg3 2592, dbg4 2720, dbg5 4241, dbg6 128, dbg7 65494

(000264942) Chg Registers: REG00 0x48, REG01 0x86, REG02 0x11, REG03 0x34, REG04 0x00, REG05 0x11, REG06 0x46, REG07 0x8D

(000264944) Chg Registers: REG08 0x23, REG09 0x44, REG0A 0x03, REG0B 0x02, REG0C 0x00, REG0D 0x14, REG0E 0x00, REG0F 0x00

(000264945) Chg Registers: REG10 0x30, REG11 0x99, REG12 0x00, REG13 0x26, REG14 0x39

(000265005) BatteryCharger::adjustChargingState - Old Charge State: 0x00 Current: 0x00 Temp: 0x30

(000266115) [DekaLogInfo] Event: 1115 (safetymonitor_task.cpp,line:601)            AC_PRESENT

(000266359) Temp (K) 302, AvePow 0, NomAvailCap 4392, FullAvailCap 4478, FullChrgCap 4285, FullChrgCapU 4327, SOH (%) 90, SOHS 1, SOCU 99

(000266361) Status 0x009A, Flags 0x0088, OpConfig 0x25F8, DesignCap 4500, TSVoltagePCNT 48

(000266365) dbg1 5400, dbg2 2980, dbg3 2592, dbg4 2720, dbg5 4241, dbg6 128, dbg7 65494

(000266366) Chg Registers: REG00 0x66, REG01 0x86, REG02 0x11, REG03 0x34, REG04 0x00, REG05 0x11, REG06 0x46, REG07 0x8D

(000266368) Chg Registers: REG08 0x23, REG09 0x44, REG0A 0x03, REG0B 0xC6, REG0C 0x02, REG0D 0x13, REG0E 0x55, REG0F 0x58

(000266369) Chg Registers: REG10 0x30, REG11 0x99, REG12 0x00, REG13 0x26, REG14 0x39

Interesting register descriptions:

REG09 0x64         BATFET_DIS

REG0A 0x03        Boost voltage

REG0B 0x02        VBUS_STAT

REG0C 0x02        Status including NTC_FAULT

REG0D 0x14        VINDPM Threshold

REG0E 0x55        VBAT

REG0F 0x58         VSYS

REG10 0x30         Thermistor

REG11 0x99         VBUS

REG12 0x00         IBAT

REG13 0x26         Input Current Limit in ICO

REG14 0x39         Config