BQ34Z100-G1: questions

Yusuke Mukuno

Part Number: BQ34Z100-G1
Other Parts Discussed in Thread: BQ34Z100, BQSTUDIO

My customer have some questions about bq34Z100-G1.

Question1

There is a wiring distance between the battery and the system board (bq34Z100 mounted).
The wiring impedance may be large.
Are there any concerns? Please tell me the solution.
Isys_max = 30 A, Z (wiring) = about 100 mohm

Question2

Could you please let me know about the calibration method at the timing of mass production?
Can I2C be used in customer PCB?

Question3

At Step 8 Run an Optimization Cycle,
Is it possible to measure multiple battery?
Is it possible to make golden file by the average?
Do not you need to consider battery variation?

Thanks

Muk

over 6 years ago

0 Bryan Kahler over 6 years ago

TI__Mastermind 25955 points

Hi Yusuke,

1) Most configurations keep the board close to the cells. Additional RF filtering may be required.

2a) Calibration may be performed during mass production via I2C commands. Often, when creating the golden image, values from 10-30 boards will be averaged together, assuming low variance.

2b) I2C may be used in the customer PCB. After the golden image has been created, it may be exported in a DFFS or BQFS format. Both of these formats contain I2C commands in the filestream format and can be easily parsed in a microcontroller to send I2C commands. For more information on how to parse these files, please refer to this app note: www.ti.com/.../slua801

3a) Yes
3b) Yes
3c) Calibration is for the device and the board, not the batteries. Ensure variance is kept low if averaging the calibration values. The battery should be of the same type and size that was matched to the chemID during engineering steps leading to the creation of the golden image. The gauge continually learns the impedance of the battery.

Sincerely,
Bryan Kahler

0 Katsuhiro Kanao over 6 years ago in reply to Bryan Kahler

TI__Intellectual 1710 points

Hi Bryan

Thank you for your support.

1) Most configurations keep the board close to the cells. Additional RF filtering may be required.
Using application is system side. This application current is high and wire becomes long.

(Isys_max = 30 A, Z (wiring) = about 100 mohm)

It is difficult to GG closes the cells.

So Voltage accuracy of bq34z100 would be different by wire impedance.

In this case, what is expected for GG accuracy ?

We guess GG would calculate wire impedance as battery impedance by learning cycle.

Wire impedance would be eliminated. Is this correct ?

2a) Calibration may be performed during mass production via I2C commands. Often, when creating the golden image, values from 10-30 boards will be averaged together, assuming low variance.
We want to know what I2C command to do calibration. And what command to store calibration data to write Data flash.

Would you please let me know which page is described in www.ti.com/.../slua801.pdf.

We could not find calibration method from it.

best regards K.Kanao

0 Bryan Kahler over 6 years ago in reply to Katsuhiro Kanao

TI__Mastermind 25955 points

Hi K. Kanao,

Even in system side configurations, the gauge may be located near the cells. Accuracy may be impacted. Please refer to CC Offset and Board offset in the datasheet.

In answers 2a and 2b above, calibration would be performed on 10-30 boards using bqStudio, averaging values, and assuming low variance using the averaged values in the golden image that would be used to program every device in production. The DFFS and BQFS filestream formats parse easily into I2C commands.

For performing calibration with i2c, and not bqStudio:

Please refer to the flowcharts in chapter 5 for this device for performing calibration via i2c commands to understand the sequence required for i2c and calibration on the host MCU: www.ti.com/.../sluubf7a.pdf

Please note: this is for a different device so some commands and values may change. Please use the sequence from this app note, but refer to the bq34z100-g1 device datasheet for commands and values.

Sincerely,
Bryan Kahler

0 Katsuhiro Kanao over 6 years ago in reply to Bryan Kahler

TI__Intellectual 1710 points

Hi Bryan

Thank you for your support.

3. Golden image file( bp.fs)
This have DataFlash and Firmware, which area of this is DF and FW ?
Is it possible to know Firmware version from this ?

4. What value of DF conifig Version/Manufactuer ?
Is this 0 (default ) ? or updated ?

best regards K.Kanao

0 Bryan Kahler over 6 years ago in reply to Katsuhiro Kanao

TI__Mastermind 25955 points

Hi K. Kanao,

3. Golden image file( bp.fs)

This have DataFlash and Firmware, which area of this is DF and FW ?
>> It contains both. You can try to isolate by using a delta of BQFS and DFFS. The format between both is the same. BQFS includes both FW and DF and DFFS is just DF.

Is it possible to know Firmware version from this ?
>> The firmware version should be known from the creation of the BQFS and/or DFFS files. In the Golden Image tab of bqStudio you are able to export SREC BQFS and DFFS at the same time. Program the device with the SREC or BQFS and then query the device for firmware version.

4. What value of DF conifig Version/Manufactuer ?
Is this 0 (default ) ? or updated ?
The current firmware version is 16 for the device (as of this writing).

Sincerely,
Bryan Kahler

0 Katsuhiro Kanao over 6 years ago in reply to Bryan Kahler

TI__Intellectual 1710 points

HI Bryan

3.Golden image file(bp.fs)

3-1 Which area of this is DF and FW

Customer strongly request, which data of bqfs is Firmware ?

>You can try to isolate by using a delta of BQFS and DFFS.

I confirmed by using diff command but result failed, these are too different.

So would you please confirm and reply it ?

Or would you please let me know which address is FW and which address is Data Flash in bq.ps?

3-2 Firmware version data from bp.fs

Customer is afraid of FW changing of bq34z100 that will be provided at mass production.

So they want to check FW version by comparing reading bq34z100 and customer's bp.fs

Would you please let me know which line could know Firmware version from bp.fs ?

4. 4. What value of DF conifig Version/Manufactuer ?

Is this 0 (default ) ? or updated ?

My question is not FW version. DF conifig Version was 0.

5. Calibration data

When customer did calibration before writing Golden image.

Is calibration data changed or kept ?

6. FW Update

When the TI change the FW. When Customer want to update it by their production line.

What data will be proved ? Is bq.fs or srec

0 Bryan Kahler over 6 years ago in reply to Katsuhiro Kanao

TI__Mastermind 25955 points

Hi Katsuhiro,

Thank you for all these great questions - due to current loading I will be able to provide a response on Friday 2019-02-08.

Sincerely,
Bryan Kahler

0 Bryan Kahler over 6 years ago in reply to Bryan Kahler

TI__Mastermind 25955 points

Hi Katsuhiro,

>> 3-1 Which area of this is DF and FW

This information is not typically disclosed without an NDA. If this matter is pressing, we can handle it offline.

>> 3-2 Firmware version data from bp.fs
The bqfs overwrites the firmware and dataflash, so the firmware shipped on the bq device from the factory will be overwritten.

As of this writing, the shipping version is 16

>> 4. 4. What value of DF conifig Version/Manufacturer

0 by default

>> 5. It is kept when exporting, but programming the device with a golden image will overwrite the calibration data too.

>> 6
Save both SREC and bqfs. Bqfs/dffs is the easiest way to program the device in production

Sincerely,
Bryan Kahler

0 Katsuhiro Kanao over 6 years ago in reply to Bryan Kahler

TI__Intellectual 1710 points

Thank you for your support.

5-1 . Voltage Calibration

When is the recommendation timing for doing Voltage calibration ?

Customer is multi cell application, so it is necessary to set voltage divider.
I thinks there are 2 timings to do set voltage divider,
One is before learning cycle to generate golden image, the other is after writing golden image at customer's production.

Which is TI recommendation ?

To get more accurate and considering diver resistance accuracy, after writing golden image seems better.

best regards K.Kanao

0 Bryan Kahler over 6 years ago in reply to Katsuhiro Kanao

TI__Mastermind 25955 points

Hi K.Kanao,

Voltage divider will need to be set for the golden image.

In production, it may also be necessary to calibrate each board for voltage divider because the components are external and may not have low ppm drift or high tolerance.

When creating the golden image, please perform the calibration procedure on 10-30 boards. If the calibration values between the boards have low variance, the values may be averaged together in the golden image.

After the golden image is created, test the results on another set of boards.

If the customer is happy with the SOC results of the golden image programmed on to 10-30 other boards without calibration, voltage calibration may not be required in production.

Sincerely,
Bryan Kahler

0 Katsuhiro Kanao over 6 years ago in reply to Bryan Kahler

TI__Intellectual 1710 points

Thank you for your supprt.

5-2 Voltage Calibration

>If the customer is happy with the SOC results of the golden image programmed on to 10-30 other boards without calibration

Would you please let me know how to check the SOC result　 is good or not.

* If the bq34z100 SCO result is 10%, how to check 　10% is good or bad ?

*How many cycle charge and discharge is needed to confirm it ?

5-3 Voltage divider resistance accuracy ?

How much accuracy is recommendation ? If customer use 1%, is it necessary board calibration on Production line ?

best regards K.Kanao

0 Bryan Kahler over 6 years ago in reply to Katsuhiro Kanao

TI__Mastermind 25955 points

Hi K. Kanao,

With respect to SOC accuracy: www.ti.com/.../slua684.pdf

For the voltage divider accuracy, it will be based upon the customer testing SOC accuracy over temperature and being happy with the results. Otherwise, I recommend calibrating voltage divider in production.

Sincerely,
Bryan Kahler

Power management

Power management forum

BQ34Z100-G1: questions