BQ78350: SOC% question

Polly Chung1

Part Number: BQ78350
Other Parts Discussed in Thread: GPCCHEM

Dear team,

Below is my customer question, could you please help give us some suggestions or work around solution? Thank you.

In my customer application, they set lithium battery charge voltage limit to 4.05V. After learning, when customer charge the battery to full, SOC% could show 100%. But when BQ78350 enter shipping mode and wakeup gauge again, gauge learning data will be reset, then that will need mapping OCV table(CHEM ID) in the gauge first to detect the battery SOC. If reset is happened, customer will found the SOC% drop to maybe 75%. We know, in general case, lithium battery charge voltage limit is >= 4.2v which might cause this problem. But is there any work around solution that can avoid SOC% drop from 100% to 75%?

GG FILE Logout is

* Device Number 1E9B

* Firmware Version 1.03

* Build Number 22

* Order Number 0

over 7 years ago

0 Bryan Kahler over 7 years ago

TI__Mastermind 25955 points

Hi Polly,

This is difficult to analyze without the gg.csv file and log file, however, have FCC, DVR, RC, Battery_Low%, and the initial value of DOD at EDV2 been adjusted as described in sections 9.1.4, 9.1.5 and 9.1.6 of the TRM ( http://www.ti.com/lit/sluuan7 )?

The bq78350 updates FCC and DOD at EDV2 when a qualified discharge occurs. The FCC should be initialized to the Design Capacity while DOD at EDV2 should be initialized to (1-Battery Low%) * 16384 where Battery_Low% = Battery_Low% / 2.56

Sincerely,
Bryan Kahler

0 JANSON CHEN over 7 years ago in reply to Bryan Kahler

Prodigy 250 points

HI,Bryan :

GG FILE used in this project please refer to the attached file,

Project information is as follows:
Battery cell: SONY US18650VTC5A (2600mA);
Chemisty ID = 0x2161
PACK combination: 13S / 4P
Initial Dod at EDV2 setting: 15936 @ Battery low 7%
Cell full charge voltage: 4.05V
Cell discharge cut-off voltage (protection): 3.0V
Application equipment load discharge voltage: 41V

Please confirm whether there is a solution to this issue.

Janson Chen

0 WM5295 over 7 years ago in reply to JANSON CHEN

TI__Guru 54715 points

Hi Janson,
On reset, the gauge will estimate SOC based on voltage, see the technical reference manual section 9.1.5. It intentionally tries to underestimate the capacity. The behavior cannot be changed.

0 JANSON CHEN over 7 years ago in reply to WM5295

Prodigy 250 points

HI,Bryan :

DOD at EDV2 and Learned full charge Capacity will be changed after completing the study,This means that I have to fill in the field with the parameters of the current study?

These parameters have been confirmed that it was written to memory, but in the upload GG FILE, not remember what you see is the initial set parameters.

The additional file is the Log file for this project during the learning process. The initial default capacity is 730, and after learning it is changed to 716. This result is similar to the result I got from PACK test equipment.

Janson chen

1-22(Learning).log

0 Bryan Kahler over 7 years ago in reply to JANSON CHEN

TI__Mastermind 25955 points

Hi Janson,

Yes, the parameters need to be filled in so the device has somewhere to start from.

How have you determined initial default capacity?

Sincerely,

Bryan Kahler

0 JANSON CHEN over 7 years ago in reply to Bryan Kahler

Prodigy 250 points

HI,Bryan :

Thank for your response, I will re-test and write the Learning result again to confirm whether this problem is prepared.

The initial default capacity I was through the Pack test equipment, and with the actual application parameters for evaluation (please refer to the attached file)

I used the following conditions and did a cyclic test to get the capacity result (about 7.2Ah)

Test conditions :

CC/CV Mode Charging (10A,1C) : 52.65V (4.05V @ Per Cell) , Cut-off Current = 100mA

Discharging (10A,1C) : Cut-off Voltage = 40.0V (3.076V @ Per Cell)

You can get results by steps 2, 4, 6, and the Status column, and by searching for keywords such as StepFinishByCut V / I / T.

Janson

13S_4P TEST DATA.xlsx

0 WM5295 over 7 years ago in reply to JANSON CHEN

TI__Guru 54715 points

Hi Janson,
Let us know if there are further questions.

0 JANSON CHEN over 7 years ago in reply to WM5295

Prodigy 250 points

HI,Bryan:

Good day ,

After I learn again, and re-enter Dod at EDV2 parameters, after the shutdown restart again, I still can not get the ASOC and RSOC before the shutdown, I will pass this test through the attached file description to you.

BTW,I try to increase the parameters of Dod at EDV2, I can get closer to the completion of the learning capacity, but the maximum value of this parameter can only be set to 16384, I try to fill in the maximum value, the resulting ASOC & RSOC About 90%.

According to the dealer's FAE instructions, Will be restarted again after shutdown, will reference Chemistry ID, the theory is correct?

But I can not get a practical application of data and curves (4.05V ~ 3.0V per Cell) .

Do you have a better suggestion?

Janson

RE-LEARNING(78350).pptx

0 Bryan Kahler over 7 years ago in reply to JANSON CHEN

TI__Mastermind 25955 points

Hi Janson,

Yes, upon reset after shutdown, it will reference the values from the Chemistry ID. The default value programming into the device is 0x1210. Please ensure the chemID programmed into the device matches your cell. This will help improve the initial estimated value prior to learning.

As WM295 posted above, the device handles initial battery capacity at device reset (as fully described in section 9.1.5 of the TRM).

Have you used the gpcchem tool to determine and set the chemID for your cells?

Sincerely,
Bryan Kahler

0 JANSON CHEN over 7 years ago in reply to Bryan Kahler

Prodigy 250 points

HI,Bryan:

In my opinion, Chemistry supported by current TI should be a parameter of Cell's standard charge and discharge characteristics. Most of the parameters in this database are likely to be the curve parameters of 4.2V ~ 2.5V (or based on the recommended application parameters of CELL manufacturers)
However, my initial choice of SONY US18650VTC5A should be the same.

As far as I know, only gpcchem tool can be used to select the approximate ID of the actual application, and I am not sure whether it can be used in my actual application (application conditions of 4.05V-3.0V). And after restarting again in Shutdown mode Provide a reasonable ASOC or RSOC.

If this method can be solved, can you suggest an EVM for the gpcchem tool test?

If this problem can not be solved, forcing us to only use the sleep function, but this will cause the battery pack can not be stored for a long time, because there is a lot of difference in power consumption Sleep & shutdown mode
The actual measurement data is as follows:
Sleep mode ~ 3-4mA
Shutdown mode ~ 300uA

Janson

0 dMax over 7 years ago in reply to JANSON CHEN

TI__Genius 14985 points

Hi Janson,
Can you please also attach your golden .senc file?
We may need to modify some settings since you will only charge to 4.05V.
To use GPCCHEM you should charge to the full voltage supported by your cell (4.2V, for example) and it will find the best match in our database. Then for CEDV gauges we might need to modify that chemID to trim off 4.05V-4.2V so that on reset or exit from shutdown it can initialize to 100% at 4.05V.
(This is not necessary for IT gauges.)

0 WM5295 over 7 years ago in reply to JANSON CHEN

TI__Guru 54715 points

Hi Janson,
The gpcchem tool should work on data and should not care what gauge the data comes from. I have not personally used it so I can't give an example. The expectation is that it would match something which would give something closer to 100% at 4.05V than the chemistry for your presently selected 2161 chemistry. I don't know of a table with parameters for full charge voltages to pick one manually.

The equation in the technical reference manual for Dod at EDV2 is wrong and pending a document update. The document gets confusing with data formats and parameter names, and the 2.56 factor is not needed here. The 16384 is a number for full, the intent is to find the number for the portion of that above EDV2. So if EDV2 is 7%, you want to find 93% of it. The calculation should be Dod at EDV2 = 16384 x (1 - .07) = 16384 x (.93) = 15237 This is different from the 15936 you calculated. This may affect how the gauge adjusts the Dod at EDV2 when it learns and how it then affects the SOC at reset.

The other adjustment mentioned in the technical reference manual section 9.1.5 is the RemCap Init Percent. It will allow a 110%. When I set this from 100 to 110 % I see an example change from 75% to 82% RSOC on reset. That won't get you from 77 to 100% but it may help if Dod at EDV2 is cutting it down.

Power management

Power management forum

BQ78350: SOC% question