accuracy issue with bq20z95 EVM

The Golden learning process is intended to be done on new battery packs. Can you please explain why you are trying to create a Golden Image file for an older battery pack? The manufacturer of the battery pack should have the original golden image file.

Hi,

We were just trying to know the accuracy level of the bq20z95,with help of bqEASY software. We could not change the update status from 04 to 06 in DataFlash. Even after completion of 2 cycles. Is it possible to improve the maxerror to 1% from 5% with old battery pack. can you please explain.

The issue here is that the Update Status of the pack is not already at 06. This indicates that the pack was released to the field without a learning cycle ever done on it. Because the pack is now aged, there is really no way to get update status to change from 04 to 06 unless you change the Qmax values to reflect the current usable capacity of the pack. To do that, you would have to have access to write the data flash.

Hi,

Find the snapshot of the error message that is displayed by bqEVSW software.What does that mean?

Regards

K.Raghu

Can you describe what you were doing in bqEVSW before you got this error? Were you using bqEasy? If so what step were you in?

Hi,

I was conducting a charge/discharge cycling using the bq20z95EVM-001 evaluation module, for a battery pack.

Is this error is a serious problem?

I am using bq20z95EVM-001 evaluation module.

I have completed one charge/discharge cycle. At that point i also completed charging the 2nd cycle as well.

Do we have to reinstall the software ?

I have not seen the error before. I would close and reopen the software and continue evaluation. If the error happens again, you will probably need to uninstall and then reinstall the software.

Hi Chase,

As you said we did cycling with initial guess of Qmax as 3000mAh, which is far less than the design capacity 5200mAh.  Even after 3 cycles as well update status value is not changing. Qmax updation is not happening. When battery pack is in realaxation mode "Voltage level OK for Qmax update(VOK)" flag in operation status register is not set. when pack is charging or discharging that flag is set. what is the problem. why VOK bit is not set. Qmax updation is not happening after 3 cycles as well.

Qmax initial value of 3000mAh is choosen as Full charge capacity for previous cycles was around 2400mAh. We did all dataflashing,clibration and cycling as mentioned  in the application book as mentioned in "Preparing Optimized Default Flash Constants for Specific Battery Types" when close Qmax is known.

When we are charging the battery for the first time CF flag in Battery mode is set, indicating Max Error is more than 100%. During that charging Cell Balancing was also not happening.

Regards,

Raghu

Please attach a log file of your learning cycle procedure. You can log using EVSW. Use a 4 second log interval. Also, please export GG files in the data flash every 300 seconds.

Hi,

We have conducted charge and discharge cycles for moli battery pack of (chemistry ICR18650J) 2s4p configuration. We followed instructions in the bqEASY and we have created the golden Dataflash image. In the process of saving golden image as suggested in bqEASY User's guide SLUU278, after performing the learning cyle. The next step is "(5c) Write previously saved .DFI binary file to the gauge to insure no cycling corruption" involves writing DFI to gauge, Does this involve flashing.  What is the saved .DFI binary file here. We have just followed these steps in the User guide till the last step and saved golden .DFI file.

After creating this .DFI the gauge could not communicate with the bqEASY. It is showing the error window as shown in the attachment.2816.snapshot_no comm.doc

 what would have gone wrong? why the evaluation module EVMbq20z95-001 is not communicating with the bqEASY?

The DFI file is basically a copy of the entire data flash. The gauge shouldn't have to go into ROM mode to program it.

As for the no communication error, It looks like something went wrong. You should still be able to communicate to the gauge after the procedure. There could be a few things here that could be causing the no comm.

First, I would try unplugging the EV2300 from the computer and reconnecting.

If that doesn't work, try sending command 08 to gauge through the PRO screen. If the device somehow went into ROM mode, this will bring it out.

If that doesn't work, perhaps the gauge lost power. Try shorting PACK+ to BAT+ momentarily. This should power the gauge back up. If that doesn't work, try checking the REG pin (pin 32). You should see 2.5V. If you don't, then it looks like the gauge was damaged somehow. This can happen if power or communication is lost during the DFI update process. If this occurs, then the gauge is permanently damaged.

Hi Chase,

I tried all the ways you told me to establish communication but it is not working.

I tried to send SMB command 08, But bqEASY says there is communication error so command can't be written to gasgauge.

I checked for the voltage level on REG pin (pin 32). It is 2.5V only. The bqEASY says that the target is not supported. Is there anyway to reset the gasgauge entirely without the help of bqEASY, as it can't communicate with the gas gauge. 

Hi chase,

Please find the snapshot of the error message displayed by the bqEASY. I neglected all the error messages and opened the software and went to bqEASY screen, than i found this error message saying " the device is having unsupported .encr file" which is required for Dataflash.3730.screenshot_unsupported_target.doc

You can power down the device physically and then power it back up. Just remove disconnect the battery and the reconnect it. Make sure to temporarily short PACK+ to BAT+, otherwise the gauge will not turn back on. If that doesn't work, try powering the gauge from the PACK+ and PACK- inputs only.

I suspect it may be with your EVSW, though. If the above doesn't work, completely uninstall EVSW and then reinstall it. Make sure to check power.ti.com for the latest version of EVSW for your device.

Hi,

We have tried all methods you suggested. But the EVM is still not communicating with the bqEVSW. What is chemistry updation? After creating the Golden Dataflash image file, We have followed steps in 11.3(preparing optimized default constants for specific battery) in application book (SLUA 404).

These steps involve updating firm ware with .senc file. For this, We have followed all the steps suggested in upadating firmware (chapter 12) in the application book.

regard,

raghu  

It sound like the gauge has been damaged, then. This can happen if the gauge loses power while updating a DFI or SENC file.

As for chemistry updating, this should have been done before creating your golden image. The correct chemistry is critical for obtaining a good learning cycle. This can be done in bqEasy.

If the device is not communicating, then you won't be able to update the firmware. You should try another board out.

Hi Chase,

What is .ENCR file? Does it contain only Default Dataflash? or does it have default chemistry file included in it.?

What default files are flashed into the new EVM before it is shipped?

Is there a default chemistry file flashed into the EVM when it is shipped?

Please  clarify my doubts

Regards,

K.Raghu

Hi Chase,

Is it possible to replace only the IC bq20z95DBT on the board?

If we replace the IC, will it work?  As we can get this IC and replace it on the board. Is it a good idea?

Will gauge work properly? Do we need any additional flashing of files before we replace the IC?

Thanks & Regards,

K.Raghu

Yes, you can change out the IC as well. You will need to rewrite your golden file to the new IC after replacing it, though. The new IC will come with default firmware.

An ENCR file is just default data flash with ID 100. ID 100 is the default chemistry that comes on all ICs.

HI,

What happens when any of the permanent flag is raised leading to permanent failure of the FETs. Can't they be fixed by writing PF key in a sealed gauge.

As once the gauge is sealed we do not have access to all commands. So this means, some how once the FET failure flag is raised the Smart battery is useless..

In the EVM once the learnign cycling is over and golden image is created with max error 1%. If we use this golden file in the production line flashing in all battery packs and install in system. Will the gauging continue to work with 1% accuracy?

Thanks & regards

K.Raghu

Permanent failures are generally used to disable a pack permanently. They are usually configured to blow a fuse. It is possible in data flash to disable the fuse blow action for any of the permanent failures if you don't want to blow the fuse.  However, the FETs will still be turned off and can only be turned back on again by sending the PF Keys to the gauge. The gauge needs to be unsealed in order to do this. There is no way to do it when the device is sealed.

The gauge will continue to work with 1% accuracy, but generally this is only at room temperature and with a similar load that you used during learning. If conditions change, such as temperature or loading, you may see accuracy decrease some, but it should remain within 3% for most conditions.

Hi Chase,

Will TI provide us the DataFlah golden image file for the gasgauge IC bq20z75?

For that what specifications do you need.? We are thinking to use moli ICR18650j battery pack in our product. Can you provide us the final golden image file which we can flash into the gasgauges in our production line.

If you can provide us the golden image file and we flash it into the gauge.. will it work with the accuracy of 1% ...?

Do you need our load characteristics as well ..to provide us DataFlash golden image file..