BQ27531-G1: New Learning Cycle

Hi Alexandru,

I believe this timer becomes set when the device is in fast charging mode, so there might be a threshold that is not being met. Can you ensure that the gauge is not in precharge mode, along with the current being greater than the charge current threshold so the gauge knows that there is charge occurring?

Regards,

Anthony Baldino

Hi Anthony,

i've investigated further my problem with the Precharge current and Safety timer and it's look that the problem occur due to the control of the charger through GasGauge.

I've diconnected on my PCB board the communication interface between Charger and GasGauge and connected directly to the PC through USB Adapter Interface. 

To control the charger i've used the bq2419x EVM-GUI software. 

First i change the precharge current from default value of 256 to 128mA and it's works when i change the battery voltage to 2.9V.

Also the satety timer expired after 5hrs of charging and stopped it.

What is your opinion?

Regards,

Alexandru Mixich

To your question from your previous post, the picture with bqstudio shows the bits from Reg8 of the charger showing "Fast Charging" is in progress.

Hi Alexandru,

I am not familiar with how the bq2419x and the EVM GUI are used, however I believe this is interesting since the precharge current and timer parameters are read/write from the gauge, so there may have been a communication issue that caused them not to be set.

However, now that these values are set within the charger, is it possible to see if the timer alarm is able to be observed from the gauge once the time expires?

Regards,

Anthony Baldino

Hi Anthony,

i followed your suggestion and reconnect the charger from PC to gasgauge to see the behaviour. 

And for precharging i found the reason why it has not working. In Reg4 of the charger there is a Bit BATLOWV which set the threshold between precharge and fast charging. When connected to gasgauge this Bit was set to 0 meaning that precharge starts when Batt voltage is below 2.8V. When testing with my Batt. simulator i need 3.0V. But to change this Bit i need to use the Advanced Command. Using DataMemory is NOT working. It shows not the content of the Reg4.

A second issue, observed during the test is that the precharge current will flow only if the voltage threshold is switched on/off. If the threshold value will be reached by changing the Batt. voltage from 3.8V for example  to 2.9V slightly the precharge current will NOT flow.

Can you tell me please why i can not set the Bit 1 in Reg4 though DataMemory location?

Concerning the safety timer, i will do similar steps next week after easter holiday.

Happy easter.

Alexandru Mixich

Hi Alexandru,

I looked through the most recent version of the firmware and saw that the Reg 04 bits are represented correctly unlike the representation from bqStudio above. Can you please tell me what firmware version is being used on your gauge currently?

Regards,

Anthony Baldino

Hi Anthony,

the FW version used in my gauge is 1.02.

Regards,

Alexandru Mixich

Hi Alexandru,

Sorry for the inconvenience, would it be possible to try and write to the Reg04 location within the Advanced Comm Tab in bqStudio to see if this is able to change the BATLOWV bit?

Regards,

Anthony Baldino 

Hi Anthony,

with Advance Comm Tab it was possible to change the state BATLOWV bit. 

However, the precharge current remain at 256mA although the value setting in the Reg3 is 128mA (see picture below). Do you have an explanation for this?

You mentioned in the previous post that the representation in the bqstudio for the Reg4 was showed correctly. Which version of the FW and bqstudio you have used? 

But the crucial issue still represent NOT stopping of charging when the safety timer elapse during GasGauge control. Do you have the possibility to test this behaviour at your place?

Regards,

Alexandru Mixich

Hi Alexandru,

Alexandru Mixich said:

However, the precharge current remain at 256mA although the value setting in the Reg3 is 128mA (see picture below). Do you have an explanation for this?

Before the gauge was reintroduced as the host, was the charger setting for the precharge current still at 128mA?

I do not believe this should be different, however can you please confirm that the Reg 03 Default in Data Memory -> Charger Info also represents the 128mA precharge current?

Alexandru Mixich said:

You mentioned in the previous post that the representation in the bqstudio for the Reg4 was showed correctly. Which version of the FW and bqstudio you have used? 

I was reading through the V1.02 firmware files. I will reach out to our firmware team regarding why these registers do not match on bqStudio.

Alexandru Mixich said:

But the crucial issue still represent NOT stopping of charging when the safety timer elapse during GasGauge control. Do you have the possibility to test this behaviour at your place?

Regarding the safety timer, I will look more into this. Just to confirm, is the EN_TIMER being disabled and enabled after changes are made?

Regards,

Anthony Baldino

Hi Anthony,

yes, before the gauge was reintroduced as the host, the charger setting for the precharge was at 128mA.

Yes, for the changes i've disabled first the timer and reenabled it after.

Regards,

Alexandru Mixich

Hi Alexandru,

I have reached out to our chargers team to get more insight into what could be causing this issue.

I'll update you with their response when received.

Regards,

Anthony Baldino

Hi Anthony,

do you have some new updates concerning my issues?

Regards,

Alexandru Mixich

Hi Alexandru,

Sorry for the delay. After talking to the charger team about this issue, below was their comments:

1) For the IPRECHG in REG03, there apparently is a 128mA offset built into the setting. Instead of using the current setting shown above of 0001b for the IPRECHG setting above, please try setting this value to 0000b.

2) They confirmed safety timer is correctly configured to 5 hrs. The timer will only count when device is in fast charge. When device transitions from precharge to fast charge the timer resets. Can you confirm that the full 5 hrs has been waited with device in fast charge phase (I.e. CC and CV)? Additionally some actions can reset the timer and some can result in the timer counting at half clock rate. These actions are stated in the BQ24192 datasheet on page 21 Once it is confirmed the timer is not being slowed down or reset customer should allow charge for > 5 hrs. The clock may not precisely trigger the timer fault at 5 hrs.

Regards,

Anthony Baldino

Hi Anthony,

1) setting the value in REG03 to 0000b helped me. Now the precharge current is set correctly.

2) The safety timer it's still not working. I attached the .log file taked for the whole period where charging was issued. The whole time the device was in fast charge phase and no actions were taken to reset the charging process.

Please take a look over the log file. Maybe you can find an explaination for this strange behaviour.

SafetyTimer_20240523_0945.log

Regards,

Alex

Hi Alex,

Thank you for the update. Just to confirm, the timer is supposed to alter the charge fault bit values in REG09 when the time expires, however these are being read in from the charger. Can you please confirm that the change in the charge fault bits are not being seen from the charger side as well?

Regards,

Anthony Baldino

Hi Anthony,

what i can confirm is that when i use the charger alone, without the gasgauge, the timer expiration is marked in REG09 through the bits 5 and 4 which are set to 1.

I'm not able to disconnect the charger from the gasgauge and continue directly the communication through a PC.

You were able to go through the .log file?

Regards,

A. Mixich

Hello,

Today's a holiday and no one is in the office. We will get back to you tomorrow.

Regards,

Nick Richards

Hi Alexandru,

I have looked through the log file to see if there is anything else that might be not allowing the safety timer to go off, however I cannot see anything upon first glance that could be causing this since these values are read from the charger.

If possible, can you please tell me if there is anything different being done between the charge cycle of the charger alone compared to the gauge? Or if there are any other parameters that could be different between the charger registers and the charger registers seen by the gauge?

Regards,

Anthony Baldino

Hi Anthony,

i do nothing different between the charge cycle of the charger alone compared to the gauge. Also the parameter i've used are the same in both cases. But maybe it is possible, like in the case with the precharge current, that the gas gauge deal with the configuration of REG5 NOT correctly or in other manner? Or maybe it suppress the message in the firmware?

Best regards,

Alexandru Mixich

Hi Alexandru,

I will reach out to our firmware team to see if there is anything that could be anything within the gauge firmware that could be holding the timer up. I will send an update when received.

Regards,

Anthony Baldino

Hi Anthony,

do you have reach out the firmware team? I did further a test with a real battery (instead of battery simulator) but with same results. The charge timer expiration is NOT shown in the Fault Register 09.

It's possible to test the behaviour at your location?

Alexandru Mixich

Hi Alexandru,

I have reached out to the firmware team, and will ping them again to get their response. After looking through the log sent earlier again, I see that the TREG bit is set throughout the entire log, which could potentially double the safety timer. Is it possible to see if this bit is set when the timer correctly goes off with the charger alone?

I do not have a bq27531EVM or the charger EVM on hand, however I will look to acquire one.

Regards,

Anthony Baldino

Hi Anthony,

i've looked also in the log file, but the bit 1 (Thermal Status) of Register 08 is NOT set to1 (TREG). Is set to 0 (NORMAL) for the entire log.

Regards,

Alexandru Mixich

Hi Alexandru,

Sorry for the confusion, you are correct. As I read more into this, I believe that something is going on with the read of REG09 from the charger, since REG08 and REG09's sole purpose is to read out from the charger. I also checked if there were any I2C faults between the gauge and charger and there is no error apparent.

REG08 and REG09 are supposed to refresh at the same 1s rate, unlike the other registers which only update when there are data changes. REG08 is able to change fine seen in the log file, however if we are able to prove that REG09 is able to change then we can take it being a communication off the list.

Would it be possible to try and produce one of the other REG09 faults, to prove that the gauge is able to report a change in the REG09 value?

Regards,

Anthony Baldino

Hi Anthony,

unfortunatelly i was not able to produce one of the REG09 faults.

I interrupted the I2C connection between charger and gas gauge but this failure was indicated in CHARGER_STATUS Reg bit 6, so the WATCHDOG_FAULT in REG09 could't be set. 

I tried to generate BAT_FAULT. I've set the charging voltage to 4.1V and connected a battery(simulator) with 4.2V. But the bit was not set.

By connecting the pin OTG of the charger to High, has interrupted the communication to bqstudio.

It seems that the REG09 faults can't be reported by the gas gauge. Is this the problem?

Regards,

Alexandru Mixich

Hi Alexandru,

Thank you for doing those test, that gives us a much better idea of what might be occurring here. Based on this, I also believe the gauge is having an issue with reporting REG09.

Just to confirm, were the faults checked through the charger to ensure that they were being triggered in REG09 as we saw earlier when the timer was going off was being represented from the charger side?

I will pass this information on to our firmware team to see if they have any insight to how to get the gauge to read this register.

Regards,

Anthony Baldino

Hi Anthony,

yes when the charger was working alone (without controll by the gasgauge) the faults were triggered in REG09. I've just tested this with the safety timer.

Regards,

Alexandru Mixich

Hi Alexandru,

Thank you for the confirmation, I have sent all of this new information over to our firmware team for them to try and find why REG09 is not being properly read by the gauge. I will send their response when received.

Regards,

Anthony Baldino

Hi Anthony,

i have an additional question regarding "Final Voltage Time". I'm just wondering at which value this parameter is set respectivily where i can change it.

Regards,

Alexandru Mixich

Hi Alexandru,

I believe that this Final Voltage Time value is set to 0s, and is not configurable by the user. This should allow for the RemCap and SOC to reach 0 once the voltage hits the Final Voltage, which should be set to the same as Term Voltage.

Regards,

Anthony Baldino

Hi Anthony,

but if the value is set to 0s what will hapen if a high load is applied at low SOC (5 - 6%) for a short time 2-3 sec., where the Voltage is already Low and could drop under the Final Voltage?  Is there not a risk that the SOC will be set to 0 to quickly, although  physicaly the Capacity is NOT 0.

Regards,

Alexandru Mixich

Hi Alexandru,

I believe the Final Volt Time is set to 0 to match the functionality of the Term Voltage, which will automatically set both Rem Cap and SOC to 0 if the measured voltage passes it. I believe this is geared more towards the protection side of the gauge, since pushing the voltage any lower consistently could cause damage to the cell. 

Regards,

Anthony Baldino

Hi Anthony,

i'm just wondering if you were able to get some clarification about setting the error bit in REG09 in case the charging time is expired.

Second, with a device i have a strange behaviour. During the charging phase the FullChargeCapacity reach the desired value (4000 - 4100mAh). 

During the discharging phase, the capacity reached is about 2200mAh more less than the value showed by FullChargeCapacity - 4070mAh.

The SOC value is showing 7% although the battery voltage is 3581 mV.

When the device is switched OFF and after a time 30 - 60min again ON, the SOC value is around 50% at voltage of 3528mV.

The stateofHealth is 2 over all phases. 

How it possible to have, at aprox. same voltage 3500mV, so big differences in the SOC value (7% and 47%)?

I change the battery to exclude the possibility of a defect, but the behaviour was the same.

I have to mention that this behaviour exclusivelly happens only at one device. Other devices works fine but have for StateOfHealth the value 3.

Regards,

Alexandru Mixich

Hey Alexandru,

Anthony is out of office today, he should get back to you next week.

Regards,

Nick Richards

Hi Alexandru,

Sorry for the delay, regarding the REG09 issue, I have not received a response so I will reach out to a different member of their team for their assistance. Can you please confirm if the [BYPASS] bit below is set?

Regarding the issue of the differentiating SOCs, would you happen to have a bqStudio log file depicting this issue? This would allow us to look into the registers of the gauge to try and find the root cause of this issue.

Regards,

Anthony Baldino

Hi Anthony,

the bit [BYPASS] ist not set  in our configuration because we decide for our application that the gas gauge should control entirely the charger bq24192.

Regards,

Alexandru Mixich

Hi Alexandru,

After discussing this issue with our firmware team, is it possible to change the [BYPASS] bit to 1 and reattempt the test? They believe this could assist in relieving this issue.

Regards,

Anthony Baldino