BQ78350: BQ78350 PF Status register question

Hi Phil,

Please see the conditions for when VIMR PF sets. Please check the settings in the configuration file for VIMR: Check Voltage and VIMR: Check Current


Regards,
Jose Couso

G'day Jose,

The VIMR settings are:

Perm. Fail  VIMR  Check Voltage         3500 mV      0x053e/2
Perm. Fail  VIMR  Check Current           10 mA      0x0540/2
Perm. Fail  VIMR  Delta Threshold        500 mV      0x0542/2
Perm. Fail  VIMR  Delta Delay              5 s       0x0544/1
Perm. Fail  VIMR  Duration               100 s       0x0545/2

But what I'd really like to know is what is the source data for the Bat Direct Voltage in the PF snapshot data.

It's not clearly documented how this is calculated. Is it the Bat voltage from the BQ76930 (which it's DS says is the sum of the cell voltages) or the voltage from the BAT pin on the BQ78350?

It doesn't match up to the sum of the reported voltages (out by ~300mV) which seems a bit. Which means??? Perhaps there was an cell update in between and this was a rapidly changing event...

The other thing is the pack is now happy. All voltages report normal. So this is some kind of transient event...

The min lifetime is also odd as it's got down to 2.5V. This happened during the event.

The pack is idle during this event, but under going vibration.

I've never seen a broken wire generate a voltage reading of a single cell voltage dropping.

Cells:   4.072   4.065   4.072   1.478   6.109   4.077   4.078   4.069   Avg:   4.075

Avg is the BAT pin voltage on the BQ78350

I've seen lots of data like this where a broken wire affect the adjacent cell readings.

Regards
Phil

Hi Phil,

The BAT pin voltage is the sum of the cell voltages. BAT pin in the BQ78350 should be the same as the highest cell voltage in the BQ76930. 

Please reference to the TIDA-00792 reference design for proper cells connections. Make sure you are using the correct RC filter components at the input of the cell pins. 

You can also reference to this application report for BQ76930 schematic connections. 

Regards,
Jose Couso

G'day Jose,

Yes I understand BAT pin should be the same. And under normal condition it is.

My question is about the data source for the

PF Status     Device Voltage   Battery Direct Voltage Register.

That is the register stored when PF event occurs.

This is not documented!!!!

This is important for understanding the fault condition.

Thanks

Phil

Hello Phil,

I looked at the FW code and the device does not have any conditions for battery voltage to trigger a PF alert. All the device does is store the last status of the Battery voltage when a PF event occurs.

It is just a register to look at when a PF alert occurs. 

Regards,
Jose Couso

G'day Jose,

So what is the source of the stored Bat Direct Voltage register?

What is the source measurement? BAT pin, or sum of cell voltages (Which seems redundant information)?

I'm still not clear on what measurement was used.

There's a couple of different "live" voltage measurement register which one is this PF snapshot using?

Thanks

Phil

Hi Phil,

Jose Couso said:

The BAT pin voltage is the sum of the cell voltages. BAT pin in the BQ78350 should be the same as the highest cell voltage in the BQ76930. 

The BQ76930 will read the differential voltage at each VCx pin of the AFE and it will add these voltages together to form BAT pin voltage. 

Phil Reid said:

There's a couple of different "live" voltage measurement register which one is this PF snapshot using?

Regards,
Jose Couso

G'day Jose,

Agreed BAT should be the same as the sum, +/- some error.

And this is what we get.

Here's our circuit which has Bat connected to he highest cell.

However if a fault occurs with the battery cell connections (eg an open circuit) this no longer always holds true.

Say we have ~4.07V on each cell. The sum of C4 & C5 is ~8.14V.

But lets say vibration or shock has broken the connection to  C4, we no longer get a reading of 4.07V on C4 & C5 but some other value. Such as ~1.478  & ~6.109  from the example above, which does not equal 8.14 is is 7.58. This can be explained I believe by considering the paths mention in SLUA749A section 10 - Random Cell Connection.

The sum no longer matches the BAT pin. However the BAT pin voltage shouldn't change.

This is why I'm interested to know the true source of the PF Bat Direct Voltage Register at eeprom offset 506 (0x01fa).

Here's some data from a battery were I initiated a PF by inducing of Over Temperature condition (decrease NTC resistor value to safely trigger the failure)

Sum of the voltage is: 28729.

I'm guessing bat direct is in units of 10mV for Battery Direct and not mV, but I could find documentation to confirm this.

The live voltage is is 2870 which is in units of 10mV for our setup.

With the sum of the cell voltages a equal 28699mV

The Avg Cell Voltage (Bat pin) is 3593mV which is in the middle of the cell voltage so looks ok.

So the PF Bat Direct Voltage in this test does not match the sum of the cell voltages exactly.

This leads me to believe it is a calculated form something.

Perhaps the Ext Avg Cell Voltage * number of cells / 10????

Or something else altogether????

I need to know to safely interpret  a PF status dump with the following

PF Status     Device Voltage   Cell  1 Voltage                       4095 mV
PF Status     Device Voltage   Cell  2 Voltage                       4094 mV     
PF Status     Device Voltage   Cell  3 Voltage                       4079 mV  
PF Status     Device Voltage   Cell  4 Voltage                       4094 mV  
PF Status     Device Voltage   Cell  5 Voltage                       4091 mV  
PF Status     Device Voltage   Cell  6 Voltage                       4090 mV   
PF Status     Device Voltage   Cell  7 Voltage                       3441 mV     
PF Status     Device Voltage   Cell  8 Voltage                       4086 mV
PF Status     Device Voltage   Battery Direct Voltage               31780 mV

I multiple the PF status register in this dump by 10. Which is an assumption...

Hi Phil,

It looks like battery direct voltage is a direct read from the AFE. Can you probe on the BAT pin of both the AFE and BQ78350? 

Phil Reid said:

I'm guessing bat direct is in units of 10mV for Battery Direct and not mV, but I could find documentation to confirm this.

IT should be by default in units of 1mV. Can you check the gain in calibration window? Maybe this is a calibration issue.

Regards,
Jose Couso

Thanks Jose,

That was the info I was after.

Disappointing, I was hoping it'd be the independent measurement via the BQ78350 BAT pin.

The pack in question is remote and very hard to disassemble safely.

Whatever happened to it is transient, and all voltages have return to nominal values.

All cells are reading normally and BQ78350 BAT corresponds to the AFE sum.

C7 on this pack has dropped as low as 2.5V with a lifetime Delta Cell Voltage of 1.6V.

All other cells are nominal.

It's perplexing as to what has happened.

As the AFE just sums the cell values the bat direct is not a great help.

I'm guessing the discrepancy in the sum of the cell and the Bat Direct can be put down to the AFE scheduler doing an update in between the cell voltage records and the BAT direct update. Suggesting a rapid event.

Regards

Phil

Hi Phil,

Thanks for confirming. Yeah, I am guessing you will need to calibrate the BAT pin of BQ78350 to match the direct BAT voltage of AFE.

Regards,
Jose Couso