BQ77915: battery doesn't charge complete

Hi Shawn,

Hope you are doing good

Thank-you for the reply and suggestions.

I have few questions regarding suggestions

• isn't CTRD and CTRC related to stack configurations of the device? 
• Will connecting the battery ground to CTRC and CTRD instead of true ground trigger any unexpected fault conditions?
•  I am sorry for the unclear schematic. R41 is Rocd resistor. Rts is a 10k NTC that is at our battery pack. Did you mean Rocd as 10kohm instead of 100kohm? From table 9.3 in the data sheet thatthye least value is for Rocd 100kohm is related to EEPROM delay options (EC table). I did not find a reference for EC  table in the data sheet. 
• R33 and R36 are 3 ohm resistors which are calculated values for the gate resistance from TI resource (https://www.ti.com/lit/an/slla385a/slla385a.pdf?ts=1594051416680&ref_url=https%253A%252F%252Fwww.google.com%252F#:~:text=The%20MOSFET%20or%20IGBT's%20datasheet,calculate%20the%20source%20inductance%20LS.&text=Determine%20when%20the%20series%20resistance,damped%20or%20critically%20damped%20performance. )
• Can you give reference for the calculation that you followed to calculate   R33 and R36 as 1k and 4.56k?

Thank-you

Warm Regards

Harini Krishna

Hi Harini,

I am doing well, thank you! Hope you are doing well also.

Yes the CTRD and CTRC are for stacking conditions. But the device will trigger fault mode if this pin goes high, that's why we suggest connecting them to a true ground. It is unlikely that BAT- will trigger high, so this is just to be safe. 

Ah, I see that R41 is the Rocd resistor. Apologies for the confusion. Those values should be fine.

R33 and R36 seem to be very low. I would suggest using a value closer to the recommended datasheet value. 3.32k should be a better solution.

Hi Shawn,

Thank-you for the reply.

Can you please give reference calculation for calculating gate resistance?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

As suggested,

• we replaced R33 and R36 values to 1k ohm and 3.32Kohm. ( did not connect CTRC  to true ground). The battery voltage is 12.15V . We tried charging the battery and the battery does not charge.  the Dschg fet is on and the Chgfet is off. We tried connecting 6A load and chgfet is still off. There is no indication of  balancing of cell voltages ( i absolutely see no condition to trigger the OV fault and as you said before, the chances of temperature faults are very remote). This seems to be very strange condition. 
• We measured the CRTC and CRTD pin voltages with respect to  GND (NOT BAT-) and found both the voltages to be 0.4V. 
•  does true ground mean the common ground other than bat-? we have other modules on the same PCB whose ground is isolated from the BMS ground (our case it is bat-) .I am attaching the PCB below
• The BMS ground (bat -) is completely isolated from the rest of the modules. Does true ground mean the common ground? is it the GNDREF or GND as shown in the schematic below?

Thank-you

Warm Regards

Harini Krishna

Hi Harini,

Have you observed the battery voltage on a scope? Sometimes we see spikes in the voltage that could cause a fault. Is there a way to get the battery voltage low enough to recover and try to charge again? Additionally, have you observed this on just one device or multiple?

As for GND, it depends on your particular board. I would just make sure that it is a true GND compared to the VDD of the device. Any voltage on these pins could cause issues. 

Hi Shawn,

Thank-you for the reply.

• we did not check the battery voltage on the scope but we did use multimeter and voltage measurement unit  to measure the battery voltage. There are no spikes in the battery voltage.
• Once we changed the R33 and R36, the connection of the load is not clearing the fault  condition. the CHG fet is still not turned on.
• We did have same problem with the same device bq7791501 where the chgfet turned off but the fault got cleared when the load was connected and even before the load was connected, it indicated voltage balancing.
• should the voltage on CRTC pin and CRTD pin measure 0V?
• When the load is connected and there is no fault condition, should both chgfet and dschg fet be turned on?
• As you said the true ground is compared to VDD, the pin VDD on bq77915 is connected to the pack+ and battery+ ( Figure 10-9 from the datasheet). so the true ground is pack- in this case? is it correct or i understood it the wrong way?I am confused a bit regarding true ground.

Thank-you

Warm regards

Harini Krishna

Hi Shawn,

thank-you for the suggestions.

based on the EVM of bq77915 we connected all the connections at bat- terminals to pack - terminal

c32 and c33, c31, c29,c41,c39  connected to  pack - instead of bat-

IC pin 9 ( VDD), pin 20, (CCFG), PIN 23 CTRD and PIN 24 CTRC are connected to pack- according to the EVM.

we tried the following test conditions and we observed the followiing:

• we removed the NTC connection and found out that there is no fault condition occurred. 
• we tried charging the battery and found out that the old condition of chgfet turning off is still persisting.
• We connected load to recover fault condition and tried charging it again and the chgfet turns off.
• We put back all the connections ( the connections shifted to pack -) to  bat- again and observed that the battery is charging without any fault condition occurring and again at 12.2V the charging stopped.  We are using bq7791500 and according to the specifications of the device, the charge voltage is 
• .We put back all the connections ( the connections shifted to pack -) to  bat- again and when we remove NTC  connections it was showing fault condition and the  battery was disconnected. once the NTC connections were restored, the device returned to normal condition. 

We have tight deadline for our project and the BMS is behaving very wired. It is always the CHGFET that is turning off.

Can we connect through call or skype?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

We see that at POR the device assumes to be in OV fault condition and will keep the charge fet off. once the cell voltages are below 4000mV for bq77915 , then the chgfet is turned on.  Till the cell voltages reach 4000mV, the chgfet will be turned off?

Is this condition triggering the chgfet to not turn on when trying to charge the battery pack which is at 12.2V?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

Thank-you for the reply

in the post 

it says that the grounds related to bq77915 should be connected to  bat-, but as you mentioned in the earlier post that the  true ground is  which is a compared to the VDD of the device. so Pack -ve  should be connected to the device ground points.

There is a big confusion when it comes to ground. Can you please tell clearly if it is bat- or pack - to which the grounds if the IC should be connected to?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

Thank-you for the clarification.

Can you please give a reference for the gate resistance calculation?

We followed the TI document to calculate the gate resistance and the calculated value was 3 ohm which is no where close to the value you suggested.

Can you please help us with this?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

Hope you are doing good.

Thank-you for the reply.

We had a condition where the chgfet was turned off. We measured the battery voltage to be 12.35V and  at  battery+ ( pack+) and pack- to be 11.83V.

I am a bit confused about charge fet and discharge fet operations.

From the datasheet i read that only if bot the fets are ON ( CHGET and DSCHGFET) the operation is normal. and at fault conditions based on the fault, either any one of the fets will be off or both the fets will be off.

in out condition, we measured the voltage at pack+ and pack-  terminals  to be 11. 83V. from the schematic given in the datasheet, i see that the battery - and the pack - are completely isolated when chget is off. how can we still read voltage at pack+ and pack- terminals?

Why is there a 500mV drop in the readings?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

Hope you are doing good.

Thank-you for the reply.

We watched a video about the cell balancing algorithm on bq77915 device. We are testing both bq7791500 and bq7791501

We have few doubts regarding the cell balancing

1. The cell balancing starts at Vstart = 3.8 V ( for both the devices  bq7791500 and bq7791501) and the threshhold is  (Vcbth- Vcbtl) is 100mV. Initially when all the cell voltages crosses Vstart , the cell balancing starts. In step 1 (Vcbtl=3.8 and Vcbth=3.9 V) if any of the cell voltage is above  Vcbth=3.9 V then that cell experiences cell balancing. does the charging process stop while cell balancing?
2.  When does the lower and upper threshold shift (Vcbtl=3.8V to Vcbtl=3.9 V  and Vcbth=3.9 V Vcbth=4 V)? is it only after the cell balancing in step 1 is completed?
3. In the video it is  mentioned that the once the cell voltages reached the Vfc threshold ( that is 4.1 V for bq7791501) the balancing will take place only if the cell reads an OV condition ( above 4.2 V for bq7791501). Did I understand it the correct way?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

Thank-you for the reply.

Hope you are doing good. 

we are using bq7791501 as BMS for  3s2p configuration li-ion NCR18650 batteries.

1. While discharging the battery we observed that at  8.7 V on the pack and 2.9V on each cell (voltage reading) the battery stops supplying the load. But according to the datasheet, the device should enter UV fault at any of the cell reading 2.8V or below. We had uv fault occurring at one of the cell voltage reading at 2.88V also. 
2. we observed one more condition where both the fets are turning off at 8.7V battery pack voltage. We have a constant heater load. the load current measured was 5A. from the data sheet it says it can be OCD1, OCD2, SCD, OCC, UTD, OTD, OW faults if both the fets are off. we anticipated it to be OCD1/ OCD2 fault  or SCD fault but we were measuring load current and it read 5A. since the battery was not charging, OCC can be ruled out.  the battery was at room temperature.  We do not see any open wire on the battery pack connections to PCB. We tried to recover the the fault by removing the load, but our application does not support disconnecting the load completely ( we have 3.3V regulator, buck converter and a 5 V regulator along with the heater load at the pack terminals which does not support disconnecting the load completely. keeping the pack terminals floating isi not supported in out application ). can you suggest any other alternative for load removal or clearing the faults where both the fet's are off and the load need to be disconnected to clear the fault?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn, 

Thank-you for the reply

Hope you are doing good.

We are using NCR18650b li-ion battery with 3s2p configuration. i am attaching the datasheet below

NCR18650B.pdf

We plotted the charge and discharge characteristics and found out that the discharge capacity and the capacity are has a lot of difference. the battery is discharging 5774mAh( the battery is not completely discharged) and charging only upto 3548mAh capacity. Please find the attached graphs for  the charge and discharge characteristics.

We are discharging the battery at 1c and charging the battery at 0.5C

Why is the  charge capacity very less than the discharge capacity?

Can you plesase help us with this?

Thank-you 

Warm Regards

Harini krishna

Hello Shawn,

Yes, it is odd that the BMS is kicking in at 2.9V and turning off Dsg FET on BQ7791501 which supposed to have a 2.8V UV fault protection. We have measured voltage directly at the PINS they are reading the same values. 

The observation is it is turning off when one of the cells is reaching at 2.88 or 2.89V situation has not changed. 

And one more observation, when we tried to measure the voltage across NTC it is showing 0V wrt pack- or common ground. Should it not give some value at 25'C? 

Our NTC value is 10K at 25'C.

We are unable to understand why it is turning off at <=2.9V instead of <=2.8V.

Appreciate any suggestions. 

Hi Shawn,

Hope you are doing good.

With bq7791501, we are using a 4.5 mohm sense resistor. 

to confirm things with you, for a 3s2p configuration, Device :bq7791501  and sense resistor value : 4.5m ohm , the following would be the thresholds

As suggested by you, we did look into table 9.5 from the data sheet and found that for discharge fet to be off, the only possible faults can be  a UV fault or CRTD disabled. as any of the cell is not reaching the UV threshold, the only possible fault would be CRTD.

As you mentioned in one of the previous posts, the CRTD and CRTC should be connected to ground.We did connect the CRTD and CRTC signals to ground. Aren't CRTD and CRTC pins related to stack configuration?

What condition may be triggering a CRTD fault in the device?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn, 

We want to measure the battery pack temperature with a adc channel. for SOC estimation of the battery. We want to use the temperature functionality of the device for measuring the temperature of the battery pack.

Can we use the NTC measurements to read the pack temperature?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn,

Hope you are doing good. 

We have CRTC and CRTD signals connected to ground.

when you say to measure signals with respect to battery ground or with respect to  pack negative?

as we already have both the pins connected to battery negative it would read 0V with respect to 

could you please tell with respect to which ground the CRTC and CRTD pin voltages need to be measured? with respect to pack - or battery -ve?

Thank-you

Warm regards

Harini Krishna

Hi Shawn,

Hope you are doing good.

Thank-you for the reply.

in the previous post you mentioned that "The VTB pin (and therefore the TS pin) are not a steady voltage, but rather a series of pulses", can you please tell what is the frequency of the pulse?

for temperature measurement using temperature functionality of the bq77915 device through ADC channel, how can we make use o the NTC measurements of the battery pack on the bq77915 device?

Thank-you

Warm Regards

Harini krishna

Hi Shawn,

Than-you for the reply.

I am unable to open the image attached by you in the previous post.

Can you please attach it again?

Thank-you

Warm Regards

Harini Krishna

Hi Shawn, 

Thank-you for the reply.

We want to measure the temperature using ADC channel for battery SOC measurement.

please refer the figure given below

we want to measure the temperature of the battery every 1 min to determine the soc. 

can the temperature of the battery be read as shown in the figure above?

as you said that the VTB and TS pulses are not steady frequency pulses, how does the bq77915 know when to make temperature measurements ?

Do you have any suggestions on how to read the battery temperature using NTC connected to bq77 device through an external ADC port?

Thank-you

Warm Regards

Harini Krishna