BQ27Z561: Facing problems while configuring fuel gauge via Micro-controller

Hi 

We will look into your question and get back to you shortly

thanks

Onyx

Hi Onyx,

In addition to above questions, for getting accurate battery data, does it require to program all files (bq.fs, df.fs,srec and gm.fs) or only one file gm.fs will be sufficient?

Any help regarding configuring bq27z561 via microcontroller would be appreciated!

Thanks

shalin.

"write operation in file : W: 16 05 12 00 00 10 00 07 C1 E5 66 19 67 60 67 9B 03 C1 9B FF C0"

"MCU is writing to slave address 0x16 with data "00 00 10 00 07 C1 E5 66 19 67 60 67 9B 03 C1 9B FF C0" on address "0x1205" (Is this operation correct?)"

No, this is not correct. 

This command writes to I2C device (slave) address 0x16, starting at I2C register 0x05 (auto-increment). The data is 0x12 0x00 0x00 0x10 etc.

About the .gm.fs file:

"Am i missing something here? Is it necessary that all the addresses mentioned in gm.fs file must match correctly to achive accurate data for battery info?"

The compare commands in the .gm.fs file will only succeed if the preceding write and delay commands are issued exactly as in the .gm.fs file. If you change anything then the gauge will likely return different data as the data that is read from the gauge is mapped from internal memory depending on the preceding commands.

"Is there any sequence i have to follow for programming bq27z561? like among these four files which file should be executed first / step by step? Any prerequisites before each files?"

If you want to program the bq27z561 in-system (and not with TI tools like SmartFlash) then you must implement a flash stream (.fs) parser that will execute the commands in the flash stream file (.gm.fs, .bq.fs or .df.fs) exactly.

There is no sequence other than that you have to create a golden image and then program it to each gauge during production. The .srec can't easily be programmed in-system (it's intended for bqStudio). The .gm.fs and .df.fs files contain data memory configuration only. The .bq.fs file contains the firmware also (so it's bigger). If you use silicon that already has the same firmware revision as the one used to extract the .gm.fs file then I would just program the .gm.fs file (this file has a FW compatibility check in the first compare command and won't program if the gauge doesn't match).

"Can i get some help to understand file format for .srec & df.fs file such that i can execute I2C write command on bq27z561 slave device?"

The .srec has the same information content as the .bq.fs file but it's not possible to parse without additional information that isn't public. The .fs files use the flash stream format. An example parser source code is in http://www.ti.com/lit/an/slua801/slua801.pdf

Hi Dominik,

Thanks for the reply!!

So, in my device, as the firmware of silicon is matching with that of gm.fs file, it does not require flashing the other three files. Is this the correct statement?

Also, if the above is correct, after programming gm.fs file to chip bq27 correctly gauge should give battery information accurately.

But I am still not getting the correct battery percentage value (SOC reading from 0x2C register).

• Does is it require any additional configuration or setting? (like design capacity, design energy, termination voltage or taper rate or these things are already there in gm.fs file?)
• Once the gm.fs file is programmed to silicon how could it be verified that it does not require gm.fs file as it is already done. Is there any method or register from which MCU can identify not to program gm.fs file to silicon again? (because programming silicon EEPROM again and again will reduce its life)

Thanks,

Shalin.

Dominik Hartl11 can you please help on above questions?

Hi Dominik Hartl11,

Thanks for the reply!!

• As per your suggestion, we have verified that generated golden image is giving correct battery soc value on BQ27z561-EVM011 EVK. So the learning cycle steps and golden image file looks correct.

• We have programmed gm.fs file to bq27z561 via MCU over custom board and I have verified by reading all the registers back that it is same as the BQ27z561-EVM011 EVK. [ To review: please have a look at the MCU readback registers and EVK readback registers. ]

Queries:

1. From our initial communication we assume that if the BQ firmware version is same, then by just writing the correct gm.fs why we are not getting the correct soc readings? 
2. Sorry to re-iterate again, Is it required to write the bq.fs file as well for getting the correct soc values even if we are getting same BQ version? [As in the EVK we are performing this steps: 1. Write the bq.fs 2. Write the gm.fs,  and we are getting the proper data with the EVK.]
   
3. From looking at the documents and with our previous communication we understood that in the bare chip (with the same BQ version as EVK) we just need to write the gm.fs and we will get the correct soc data. Please correct us.

gas_gauging_data_flash.gg.csv

Hi Dominik Hartl11,

Is there any other step which is required to get accurate result? Or please suggest if any system readings or environment details is required for review purposes.

I have followed the method mentioned in  http://www.ti.com/lit/an/slua801/slua801.pdf

If you get accurate results with bqStudio and the EVM and if you have verified that the configuration is the same on your system then the gauge will produce accurate results in the sense that the gauge will behave the same way as on the EVM.

So if you don't get good results there seems to be still something wrong with the configuration on the system board. I typically try to hook up the TI USB I2C interface to the system board and try to use bqStudio to check if the configuration is indeed correct. If you have access to GND, SCL and SDA (and can configure your I2C host uC to disable its I2C interface temporarily) then this is the best way to verify if the gauge was truly programmed correctly.

Hi Dominik Hartl11,

I did setup with my system board chip to EVK board using SCL,SDA,GND I2C lines and connected with EVK board on J7 as shown in below image.

But i was not able to read chip using bqstudio software. My system is having bq27z561 and battery connected to it, and i also disabled I2C interface between MCU and bq27z561.

Is there any other hardware connection required so that i can read my system chip data using bqstudio?