BQ40Z50-R2: Application issues

Hello Ian,

Can you also attach a log file from the gauge showing this behavior.

Regards,

Adrian

Hi，

I collected the logs recorded by Battery Management Studio. The operation steps are to use a 6A discharge current to stop at around 15% when the battery is fully charged, let it stand for about 3 hours, and then continue using 6A discharge. Upon checking the logs, I found that the voltage reached the cutoff voltage of 2700mv at 9% SOC, triggering a 0 jump. I analyzed three cases of batteries and found that one common feature is that the battery level is low (within 20%) and this phenomenon occurs when the temperature returns to room temperature before discharging. It is likely that the impedance of the battery cell changes after the temperature changes, and the estimated impedance is inaccurate. I have searched for relevant information about this, please help me answer a few questions:
1. Please confirm the reason for the battery jump based on the provided log
2. I have completed the learning cycle, can I use GPCRB to optimize the impedance after temperature changes?
3. The GPCRB tool needs to collect data at room temperature and low temperature separately. Is the static stage after low-temperature discharge placed in a low-temperature box or in a room temperature environment? This is not explained by the GPCRB tool.
4. Based on the phenomenon I described, is it recommended to use GPCRB tool for optimization? Or do you need to collect relevant logs for further analysis?

jump_log.csv

Hi，

I optimized it using the GPCRA0 tool here. The collected data was fully charged at room temperature and left to stand for 2 hours, then discharged with a current of 6A and left to stand for more than 5 hours. The gg_out.csv returned by the GPCRA0 tool was written into BQ40Z50R2, and I tried to discharge it again until the battery level dropped to 15% and the temperature dropped back to room temperature. When I discharged again, I found that the SOC still skipped 0. According to the Ra value you replied, the estimated internal resistance is high. Theoretically, a high internal resistance should result in faster capacity consumption estimation. However, from the SOC curve, the SOC change rate is not high. The attachment data room.log.csv is the data I collected for the GPCRA0 tool, and Jump_1. csv is the data I imported into the optimized gg.csv file of the GPCRA0 tool and tested again, which showed a jump of 0. Please help me analyze the cause of the anomaly again. If the Ra value is still abnormal, may the data I collected be inaccurate or the collected temperature need to be changed. A simple analysis of the characteristics of the battery cell is that when the low battery returns to room temperature, the internal resistance increases, causing a rapid voltage drop and achieving a cutoff jump. We hope to optimize this problem through BQ40Z50R2.

room_log.csv

Jump_1.csv

Hello Ian,

Can you please share the current gg file from the gauge?

Regards,

Adrian

Hi，

The gg_out.gg.csv file is written, and the Jump.gg.csv file is after testing the SOC jump. Please help analyze it. Thank you!

Jump.gg.csv

8623.gg_out.gg.csv

Hello Ian,

Please allow me some time to review the data.

Regards,

Adrian

Hi，

Okay, please let me know if you have any analysis results or need other relevant information for analysis. Thank you！

Hello Ian,

As I review the data, can you run 3 more charge, relax, discharge cycles and collect the data for that. I want to see if accuracy improves over multiple cycles as the gauge will continue to learn and adjust over time.

Regards,

Adrian

Hi，

I want to know if for the discharge process of the cyclic operation, we need to keep it at about 15% soc until the temperature is stable and then discharge, or just discharge it directly to the cut-off voltage. At present, in the test of fully charging, leaving it idle, and discharging to the cut-off voltage, the soc change in this process is normal and the accuracy is right. The attachment is the log data for one cycle. If you need to continue the cycle, please help me confirm the cycle steps and the charging and discharging current. Thanks!

3162.cycle.log

cycle_1.gg.csv

Hello Ian,

This is good news that the SOC is being tracked properly. Now, I wanted to ask what is the normal current load that will used in the end application? Will the end application have a constant or close to constant discharge until terminate voltage?

Regards,

Adrian

Hi,

My application field is for drones. The stable discharge current is 5 to 6 amps. The maximum fluctuation can reach a peak of 15 amps. In actual use, it won't be discharged to the cut-off voltage because 5% of the power is reserved. The current at the end of the flight gradually drops from 6 amps to 700 milliamps.

Hello Ian,

I would recommend simulating the same load profile for the drone and gather log files from the gauge to see if the gauge can accurately gauge your application.

Regards,

Adrian

Hi,

I use a 6A discharge and then let it stand still to simulate the actual test process of a drone. A really complete estimate of the flight power is accurate (from 100% to 0%). But we'll do some application tests for related situations. At present, this test is to hover and fly until the battery is at a low voltage (about 15%), and then let it stand still for 2 hours before flying and testing again. During the test, we found that the battery power jumped to 0. Later, we found that after standing still at a low voltage and then discharging, the end voltage would drop faster than during non-stop flight. I suspect it's because the temperature got lower and the internal resistance of the battery got larger. Because compared to during flight, the temperature dropped by 20°C after standing still. So when discharging again, the voltage quickly dropped to the cut-off voltage, causing the 0 jump. Do you have any other variables or optimization methods that I should consider? Thanks.

Hello Ian,

During the test, do you see the temperature reported from the gauge increase to larger than 40C? If so, then I would recommend using the GPCRB to optimize for the high temp. https://www.ti.com/tool/GPCRB

Regards,

Adrian

Hi，

How to optimize high temperature with GPCRB tool? According to the introduction, it collects charging, discharging, and static data at room temperature and low temperature. Please provide the operating steps or documentation for optimizing high temperature. Thank you

Hello Ian,

Here is the document for GPCRB: https://www.ti.com/lit/ug/sluubd0/sluubd0.pdf

Regards,

Adrian