BQ78PL114 IGRR failure

Look at the description of this safety feature in the SLUU330b Technical Reference Manual.

This could be indicating a problem with your pack, like a broken weld strap.

Relearn/Initialize should clear safety counters.

I have read the technical reference manual, and it does not describe adequately what IGRR is, or how it is calculated.

What is it a ratio of? Is it the ratio of impedance growth between cells?

What are the units for the IGR / IGRR fail limits? These were left out of the table towards the bottom of the technical reference manual.

In the IGRR section of the technical reference manual, it mentions that the default impedance for each cell is a user-settable parameter, but where can this parameter be set?

We are using high quality cells with rated discharge rates of 45C and internal resistances in the order of 1mR per cell, could such a low resistance be causing issues with impedance growth calculations?

Below is the .aux file we have generated for our cells. Are the NDI and NDI initial numbers within an acceptable range?

On a different .aux file run, the NDI and NDI initial were both '2'. There does not seem to be much resolution on these values, could that be causing issues with impedance growth calculations?

Also, cell one always has a Qmax lower than the other two cells as well as significantly different Ra values. Is this an artefact of the BQ being powered by cell 1 and does this have any detrimental effects on impedance or any other calculations? Is there a solution for this other than to use the 'generic' setting when generating .aux files?

# bq78PL114 auxillary chemistry data for chemistry type  146
# aux_gui version 2.00 - Actual
# based on H:\Program Files\Texas Instruments\bqWizard 3\Data Logging Files\take1.csv log file
# Generated for 3S1P pack
# Generated : 15/12/2010 9:55:41 AM
bqWizard # Application
2.5 # minimum BqWizard  version
1 # # of files
15 # File #
AuxChem # File Type
2 # ChunkSize
150 # Total bytes
##############################################
# Comments to be included in the .aux file
##############################################
    1 # Aux_Chem serial
  109 # Tau10
 2930 # NDI_LowT
 3030 # NDI_HighT
   15 # NDI_SOC
    0 # Spare0
   32 # NDI_GAIN
    0 # Spare1
    0 # Spare2
    0 # Spare3
    0 # Spare4
    0 # Spare5
########### cell 1 ##################
    5 # NDI_Initial
    5 # NDI
 4740 # QMax
 2367 # QRem
    0 # Rzero
    0 # Spare
   19 # Ra[ 0]
   24 # Ra[ 1]
   40 # Ra[ 2]
   41 # Ra[ 3]
   25 # Ra[ 4]
   22 # Ra[ 5]
   36 # Ra[ 6]
   35 # Ra[ 7]
   24 # Ra[ 8]
   18 # Ra[ 9]
   12 # Ra[10]
    8 # Ra[11]
   10 # Ra[12]
   17 # Ra[13]
   30 # Ra[14]
########### cell 2 ##################
    5 # NDI_Initial
    5 # NDI
 4990 # QMax
 2495 # QRem
    0 # Rzero
    0 # Spare
   18 # Ra[ 0]
   23 # Ra[ 1]
   36 # Ra[ 2]
   41 # Ra[ 3]
   27 # Ra[ 4]
   22 # Ra[ 5]
   36 # Ra[ 6]
   34 # Ra[ 7]
   23 # Ra[ 8]
   18 # Ra[ 9]
   13 # Ra[10]
   11 # Ra[11]
   13 # Ra[12]
   22 # Ra[13]
   39 # Ra[14]
########### cell 3 ##################
    5 # NDI_Initial
    5 # NDI
 4960 # QMax
 2478 # QRem
    0 # Rzero
    0 # Spare
   18 # Ra[ 0]
   23 # Ra[ 1]
   39 # Ra[ 2]
   41 # Ra[ 3]
   27 # Ra[ 4]
   22 # Ra[ 5]
   37 # Ra[ 6]
   34 # Ra[ 7]
   23 # Ra[ 8]
   18 # Ra[ 9]
   11 # Ra[10]
   12 # Ra[11]
   14 # Ra[12]
   24 # Ra[13]
   42 # Ra[14]

Does TI have any answers to these questions?

We are on a timeline and have products going out the door that could be compromised.

Here is an updated description of IGR and IGRR.

3.12 Impedance Growth Rate (IGR)

This condition is event-driven and not evaluated on a regular interval.  The test checks for abnormal impedance increases in each cell, independent of the other cells.

The following user-defined parameters govern the behavior of this rule:

(a)   Current Delta: Set in units of mA

(b)   IGR Fail Count: Set in Integer units. Setting to zero disables the function.

(c)    IGR Limit: Set in units of Percentage.  This limit specifies the increase in Normalized Dynamic Impedance from its Initial Value as defined in the Auxiliary Chemistry File (.aux). Setting to a value of 200 would allow increases in Impedance up to 2 times the original value.

(d)   Normalized Dynamic Impedance SOC: Set in units of Percentage.

Activation criteria/behavior:

When the relative state-of-charge (RSOC) is greater than Normalized Dynamic

Impedance SOC and a current step greater than Current Delta has occurred, the cell impedance growth rate is calculated for each series-cell element. Rate = NDI at Present / NDI Initial.  If the calculated cell impedance growth rate is greater than IGR Limit, then the cell’s IGR counter is incremented.  If it is less, the cell’s IGR counter is decremented, without going below zero.   If the cell impedance growth-rate counter exceeds the IGR Fail Count, then the following actions are taken:

(a)   All MOSFETs are opened.

(b)   The fault is logged into nonvolatile memory.

(c)    The IGR Status Flag (Permanent Disable Status Register Bit 14) for this condition is set.

(d)   SBS:BatteryStatus(0x16):[TDA] Terminate Discharge Alarm bit is set.

(e)   SBS:BatteryStatus(0x16):[TCA] Terminate Charge Alarm bit is set.

(f)     SBS:ChargingVoltage(0x15) is set to 0.

(g)   SBS:ChargingCurrent(0x14) is set to 0.

(h)   Cell balancing is discontinued.

(i)     The SPROT pin is set and the System Fail Flag (Safety Flags Register Bit 15) is set.

3.13 Impedance Growth Rate Ratio (IGRR)

This condition is event-driven and not evaluated on a regular interval.  The test checks for impedance growth of each cell as compared to the average NDI of the pack.

The following user-defined parameters govern the behavior of this rule:

(a)   Current Delta: Set in units of mA.

(b)   IGR Ratio Fail Count: Set in Integer units. Setting to zero disables the function.

(c)    IGR Ratio Limit: Set in Percentage units.  This limit specifies the increase in the Normalized Dynamic Impedance of each cell from the average NDI of the cell pack.  Setting to a value of 200 would allow increases in Impedance up to 2 times above the pack average.

(d)   Normalized Dynamic Impedance SOC : Set in units of Percentage.

Activation criteria/behavior:

When the relative state-of-charge (RSOC) is greater than Normalized Dynamic

Impedance SOC and a current step greater than Current Delta has occurred, the cell impedance growth rate ratio is calculated for each series-cell element.  The ratio = (NDI of Cell / Average NDI of Pack). If the cell’s IGRR is greater than the IGR Ratio Limit, then the cell’s IGRR counter is incremented.  If the cell’s IGRR is less than the IGR Ratio Limit, then the cell’s IGRR counter is decremented, without going below zero.  If the cell’s IGRR counter exceeds the IGR Ratio Fail Count, then the following actions are taken:

(a)   All MOSFETs are opened.

(b)   The fault is logged into nonvolatile memory.

(c)    The IGRR Status Flag (Permanent Disable Status Register Bit 13) for this condition is set.

PLEASE BE ADVISED:  This is an advanced feature that requires prudent use.  Activation of either of these rules causes a Permanent Disabling of the Pack (Fuse activated).  The BMS designer should have a thorough understanding of the acceptable limits of cell impedance growth for the cells being used.  If you are unsure of how your cells will perform over time or if you are unsure of acceptable performance limits, it is recommended that you strongly consider disabling these two rules by setting their respective counters to zero.

Thank you very much Jim for that detailed response. That is a much better description of IGR and IGRR, and clears up many of the queries we had.

As I mentioned before, the batteries we are using have quoted impedance in the order of 1-2mR per cell and our aux file generation runs have returned NDI's as low as 2 and up to 5.

Is the resolution and inaccuracies of the NDI calculation creating rounding errors and false triggering of the IGR and IGRR counters?

Do you have any recommendations for using the bq78pl11x with very low impedance cells such as ours?

Does TI have any recommendations when using very low impedance cells with the BQ78PL11x?