BQ40Z80EVM-020: Trouble with finish learning cycle, 'Update Status" field doesn`t change

Hello,

The update status could be not changing from 0x04 to 0x05 due to the Qmax update not being taken during the relax period after charge. This could be caused by the cell truly not being relaxed during this time.

If possible, can we please receive a log file of the gauge during this situation of the update status not changing?

Regards,

Anthony Baldino

Everything was done according to the instructions and the VOK register was cleared, and the REST register set the value, but the Update Status did not change, it relax for 2 hours and i even tried leaving it to relax for a longer period. There are no logs now, but if necessary, I can repeat the entire procedure. Please specify which period of logs you are interested in.

Hello,

Understood, thank you for clarifying.

If it is not possible to receive the log for the full learning cycle, then a log of the charge cycle and the relax period afterwards will give us some insight into why the update status is not changing. The log will allow us to look deeper into the registers to what is occurring in the gauge.

Regards,

Anthony Baldino

I collected the log while charging, the result is the same VOK and RDIS are changed, Update Status is not changed. log_file

Hello,

Thank you for sharing.

Looking at this log file, it seems like the conditions are not being met since the voltage during relax is not completely stable. From Sample 1595 to 3490, there is a slight negative current that could be affecting the OCV measurement taken during this time.

Can you please tell me what could be causing this negative current?

Regards,

Anthony Baldino

The negative current was caused by the power supply, which was turned off after charging but connected by wires, repeated the cycle with the power supply wires disconnected after charging was complete. new_log_file. (Update status don`t changed again)

Thanks for helping.

Hello,

Looking at this new log, the RDIS bit never is set.

The RDIS bit is initially set before the initial discharge seen below:

Regards,

Anthony Baldino

New log , checked that the RDIS was set before discharging and charging. What do you use to view logs on charts?

Hello,

In this log, VOK is not set when the charge begins.

If possible, can you please attempt the learning cycle with the discharge->relax 5 hour -> charge -> relax 2 hour process? I believe it could mitigate some of these issues being seen.

Also, I import the log files into Excel to create the graphs and check the data.

Regards,

Anthony Baldino

The problem with recording full cycle logs is that this equipment is located in the office and the working day is 8 hours (Discharging 2 h + Relax 5 h + Charging 2 h + Relax 2 h). If, after discharging, I leave the device to relax not for 5 hours but all night and then proceed to charging, will such a log be suitable?

Hello,

That method should work fine. However, when the initial discharge is started, please make sure the VOK, RDIS, and QEN bits are being set and being cleared during the long relax period. This will save time rather then completing the entire learning cycle and realizing they had not been set since the beginning.

Below is the normal process of the bits being set during the initial discharge:

VOK should also become set when either a charge or discharge has been observed.

Regards,

Anthony Baldino

New log for the cycle  discharge->relax all night -> charge -> relax 2 hour. The VOK and RDIS registers changed during the process but the Update Status did not change

Hi Artem,

Thank you for sharing. I believe this is being caused by the VOK not being set again during the initial discharge of the cell. RDIS seems to be acting accordingly through the process.

If possible, can you please reattempt the first set of instructions for the learning cycle discharge seen below (not the entire learning cycle, just these instructions and a little time discharging) to see if the VOK becomes set? 

Regards,

Anthony Baldino

Thanks Anthony,
So as not to waste collecting logs again, I want to clarify about VOK/RDIS.
- Before discharging it is necessary that VOK and RDIS are set
- To do this, you need to first call GAUGE_EN and then RESET
The problem is that after GAUGE_EN I have: QEN=1 VOK=1 RDIS=0, and after RESET QEN=1 VOK=0 RDIS=1 and it turns out that VOK remains unset. How then to set VOK?

Hi Artem,

RDIS should become set after the RESET is sent before the discharge. However, VOK should set by itself when a discharge or charge is observed. Normally, this value should be set right at the beginning of the discharge, however we have observed in previous cases that it can sometimes set later during the discharge.

This is an interesting case because in the log sent above, VOK is setting for the later charge cycle, but not the initial discharge cycle. I would like to confirm that it is able to set upon discharge.

Regards,

Anthony Baldino

If installing a VOC at the beginning of charging is normal, then what is wrong in the learning cycle? Why is Update Status not updated and what logs are needed?

Hi Artem,

Since VOK is not set from the initial discharge, this is inhibiting the toggling of the Qmax bit, which represents if a valid Qmax update has occurred. In the sent log4, the Qmax bit is low for the entirety of the cycle, which I believe is why update status is not changing.

If possible, can you please take a log of the gauge in a relaxed state, and then complete a discharge cycle on it. This will show us if VOK not setting is a reoccurring issue.

Regards,

Anthony Baldino

Log measurements during relax and subsequent discharge cycle. LOG

Thanks for helping

Hi Artem,

Thank you for creating these measurements. In this log, the VOK bit does set correctly at the beginning of the discharge seen below:

As you can also see from this log, RDIS is setting correctly in the beginning, then clearing quickly. I believe that this is because RDIS is set during the relax before the discharge, when it is supposed to be set right before the discharge is completed.

When the entire learning cycle is completed next, please ensure that the VOK and RDIS are set through the initial discharge, then cleared in the relax after.

If this is done correctly, then I believe Update Status should act accordingly.

Regards,

Anthony Baldino

After GAUGE_EN the QEN and VOK bits are set. After the reset, RDIS is installed and VOK is cleared. When discharging begins, VOK is not set back. What reasons could there be for this? Are there any additional documents besides the manual for the learning cycle and technical reference?

Hi Artem,

What is the update status when this occurred? Can you send the .log file?

There is also a bq40z80 specific manual for the learning cycle which can be found below:

https://www.ti.com/lit/pdf/slua848 

Regards,

Anthony Baldino

Logs. Yesterday I called the GAUGE_EN command and then called the RESET command, this set the QEN, RDIS bits. When I started to discharge the device, the VOC bit was not set and therefore I did not record the log. Today, when turned on pc, the VOC bit was set, but the RDIS bit was not. I continued to discharge the device and the RDIS bit was set, after which I began recording a log. 0How it works? and has Qmax been updated for the subsequent UPDATE STATUS update?

Hi Artem,

Artem Muzykin said:

When I started to discharge the device, the VOC bit was not set and therefore I did not record the log.

It is possible that the VOK set at some point during the discharge. As long as it sets while the discharge is occurring, the learning cycle should function as normal for this bit.

Artem Muzykin said:

Today, when turned on pc, the VOC bit was set, but the RDIS bit was not.

Was the gauge in discharge or relax when the PC was turned on?

The sent logs seemed to have VOK and RDIS in the appropriate states. Can you please tell me how the ChemID was chosen for this learning cycle?

Artem Muzykin said:

has Qmax been updated for the subsequent UPDATE STATUS update?

Update Status during both logs is still at 4, meaning the Qmax update has not occurred yet.

Regards,

Anthony Baldino

Anthony said:

It is possible that the VOK set at some point during the discharge. As long as it sets while the discharge is occurring, the learning cycle should function as normal for this bit.

I don’t understand the point that it may not set the VOC throughout the entire discharge cycle, but after I started the PC the next day and the battery was not completely discharged, when the discharge began, the bit was setted immediately.

Anthony said:

Was the gauge in discharge or relax when the PC was turned on?

After turning on the PC, VOC bit was setted initially, when I continued the discharge the RDIS was setted. 

Anthony said:

Can you please tell me how the ChemID was chosen for this learning cycle?

2129 ChemID was chosen as the most similar for this battery. Later the battery will be changed, now I need to understand how to carry out and complete the learning cycle.

Thanks for helping

Hi Artem,

When the PC was off, was the cell still discharging or was it in a relax state? This could explain why the VOK bit becomes set when the PC is turned on and discharge is continued.

Regarding the RDIS, this bit should only become set from the reset before the initial discharge. When this chemID was chosen, was the GPCCHEM tool found below used to get this ID?

Regards,

Anthony Baldino