BQ40Z50: offset calibration

Hello Brian,

Could you send some screen shots of the commands you're sending and receiving? It looks like you have the right byte to get the counter. Byte 3 indicates the mode you're in, check 11.2 in the TRM.

The error shouldn't be that high with calibration, what procedure did you follow for calibration?

Sincerely,

Wyatt Keller

Hi Wyatt,

Thanks for responding.

I am following the document SLUA734A to implement calibration of the BQ40z50.

The calibration routine outline

• Check if Sealed
• Check Fet Control if enabled disable it
• Read ADC values
  
  • Enable Calibration mode
    
    • Check manufacturing status bit 7 <high byte> CAL_TEST
    • wait for CAL_TEST set
  • Set ADC Mode with internal sense resistor short (0xf082)
  • Loop for 64 ADC samples
  • Average sample readings (0x4010)
  • Read in number of CC offset samples (64 default)
  • Write CC offset

Attached are the command strings I am reading and sending.  Let me know if you see something that I have out of sequence or need to change or if you have questions.

Thank you for taking a look at this.

Regards,

BrianCal_Comm_Transaction.docx

Hi Kang,

I suppose I can do that. 

Is the documentation for calibration procedure wrong?

Regards,

Brian

Hi Kang,

Ok I understand.  I am not sure how to start the board cc offset processes using bq studio. 

Is there a document that details the steps in bq studio?

Regards,

Brian

Hi Wyatt,

It looks like the CC offset option is available with a bqProdcution.  I am not sure what that is.

Regards,

Brian

Hi All,

What are the next steps to get the information you need?  We have several packs that are failing during calibration.

Regards,

Brian

Hi Shirish;

It looks like this does the cc auto offset not the CC offset and Board offset.

Is that correct?

Regards,

Brian

Hi Brian,

You are right. It does CC Offset which updates cc auto offset parameter. There is no feature to do board offset.

Board offset is not required unless a board offset current is observed per section 2.4.2

Board offset was not supported in bqStudio because it is not a normal use case.

Hi Shirish,

I need to do board offset for our product.  I am following the method described in SLUA734A.  When the sense resistor is shorted internally we see in some cases large offsets.  Can you explain to me why this is happening?

Can you provide some insight as to why we are seeing this?

Regards,

Brian

Hi Brian

Can you check if all layout guidelines have been followed?. The datasheet has layout guidelines in section 11

Hi Shirish,

We looked through the layout and it follows the TI guidelines.  I don't understand why the measurement with the sense resistor shorted seems to be biased in one direction.  When I make the measurements with the sense resistor unshorted (0xf081) the readings are centered around zero.

Regards,

Brian

Hi Brian,

It is unusual to have an offset current of 10mA. Do you see similar results when using the TI evaluation board?

www.ti.com/.../BQ40Z50EVM-561

Switching in/out the TI evaluation board will help isolate the reason for the offset. We have not seen offsets when using the TI evaluation board.

Regards

Hi Shirish,

Great idea; i'll try running offset cal on the eval board. 

Thanks,

Brian

Hi Brian,

The eval board is not expected to show an offset.

Regards

Hi Shirish,

I ran the calibration with the bq40z50evm-561.  

The value for CC Offset ended up being -122 which seems large.  The values are all biased in the negative direction.  

The value for Board Offset  is 132.  

When I run the CC Auto Offset the value returned is -123.  

I don't understand why the shorted voltage measurement (CC Offset) seems to be off.  Are the CC Offset and CC Auto Offset using the same method of shorting the sense resistor internally?

Regards,

Brian

Hi Brian,

The EVM should not show much of an offset. Please use bqStudio to calibrate and check the data memory values after calibration.

Regards

Hi Shirish,

Attached are the results of the current calibration performed on the bq40z50-evm.  It shows the before and after screen of the Calibration Data Memory.  

The current gain was adjusted from the default to 1.001.  Not surprisingly the board offset, CC Offset and CC Auto Offset did not change.

Can someone try verifying the procedure for offset calibration in SLUA734A? 

What would cause the offset measurement when internal short is applied to drift from zero?  

Are the parts that are returning large offset values bad?

CurrentCalResults.docx

Regards,

Brian

Thank you Brian. These results verify that the offset is large. How many parts have you seen exhibiting high offsets? I am checking with the hardware team on what could cause it.

Hi Shirish,

We are seeing about 30% of our devices with this problem.  Were you able to get feedback from your hardware group?

Thanks,

Brian

Hi Brian,

Thanks for providing additional details. The hardware group indicated that they might need more information. I wilI ping them again.

Regards

Hi Shirish,

Just checking in to see if there had been any progress.  I am starting to get a lot of attention on this as we have quite a few packs that on hold for resolution of this issue.

Thanks,

Brian

Hi Brian,

Do you happen to have the Lot # / LTC information? This will allow us to get a faster response since it will narrow down the possibilities.

How to find LTC (Lot Trace Code)

Quote from hardware team response

"Lot Trace Code is on the topside marking of the device.  Or, if they have picture of the label on the shipping reel or box that should work too."

Hardware team wants to know LTC’s with large offset as well as LTCs with good offset.

Hi Shirish,

Attached is a photo of the 40z50 from our build.

Regards,

Brian

Hi Brian,

Thanks. Is this picture of the board showing high offset? It would help if you can also send a picture of one that has low offset.

Regards

Hi Shirish,

I will get you a picture of both.  I don't think I can do that today however.

Regards,

Brian

Hi Brian,

No problem. Can you confirm that the picture posted earlier was showing high offset or low offset?

Regards,

Hi Shirish,

This board had not been used.  It is from a lot of PCB's that have been put aside because of this issue.  I have some boards available that have acceptable offset and that have failed.  I can send you of photo of both.

Regards,

Brian

Hi Shirish,

Attached is a photo with the BQ40z50.  

This photo is of a battery that passed offset calibration

Hi Shirish,

Attached is a photo from a board that failed our offset calibration.

Thank you Brian. I have sent this to the hardware team and will let you know when i hear from them.

Hi Shirish,

I am starting to get more attention from production regarding the offset problems.  The most recent production run had a 40% drop out problem.  Any help would be much appreciated.

Regards,

Brian

Hi Brian,

Thanks for reminding. I have requested a schedule from the hardware team.

Regards

Hi Brian,

This is the response from the hardware team

"

I looked into the lot vs other lots run around the same time. I didn’t see any statistically significant difference in any of the tests.

"

Hi Brian,

If you would like to investigate this further, we can open a case. That will mean a debug effort and will involve getting more details.

We will need to take that up offline

Regards