• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 3 answers
  • Subscribers 64 subscribers
  • Views 183 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ27411-G1: Recommended range for max DOD error

Fred Archer
Prodigy 60 points
Part Number: BQ27411-G1
Other Parts Discussed in Thread: BQ27426

I received a GPC report, as follows:

Device / Family #1
Generic Chem ID Device/ Voltage/ Chemistry max DOD error, %
354 bq27411-G1C: 4.35V LiCoO2 1.9
3142 bq27421-G1D: 4.4V LiCoO2 5.33
128 bq27421-G1A: 4.2V LiCoO2 10.9
312 bq27421-G1B: 4.3V LiCoO2 14.49
Best generic ID 354
Max. deviations for best generic ID is within recommended range. Please chose this ID in your device configuration or device family.

I cannot simply change to the bq27411 P/N as I need a system side fuel gauge.

The question is whether the bq27421-G1D variant with a 5.33% max DOD error would be "within recommended range".

Thanks,

Fred

  • 0
    TI__Guru** 102885 points

    Hello Fred,

    This depends on the accuracy you need from the gauge. If you're okay with higher SOC errors then this is no problem. It would be hard to give a range, but the usually we like to see the max DOD error less than 3%. I think 5% max DOD error will still give you good gauging accuracy.

    Sincerely,

    Wyatt Keller

  • 0
    Prodigy 60 points

    Hello Wyatt, Thanks for the quick response. 

    AT the moment we have [312 bq27421-G1B: 4.3V LiCoO2 14.49] variant fitted on our Hardware. 

    We get occasionally the Fuel Gauge reported SOC value jumping immediately from a value of lets say 70% to 0% after high momentary load. We know in this case the battery is ~70%.

    In order to try to eliminate this we are going through the application notes SLUA903 and firstly slva725a. 

    Steady-state Gauging accuracy itself is not our highest priority here. If we were out by 5%-10% we could work around this. The issue we are trying to solve is these immediate jumps to 0% 

    1) Do you think the non-ideal CHEM ID and lack of Golden Image could lead to these immediate jumps to 0% or do we think there is another cause.

    2) If so , if we were to proceed with  bq27421-G1B: 4.3V LiCoO2 14.49!!, and perform  SLUA903 on this relatively high error CHEM ID, would this solve these SOC jumps?

    Due to the chip availability problems, I think we are stuck with bq27421-G1B for a while, but the medium term solution, I am considering is:

    a) 3142 bq27421-G1D: 4.4V LiCoO2 5.33 - Simple BOM Change

    or

    b) 3230 bq27426: (default) 4.35V LiCoO2 1.76 -  AND make PCB change to allow bridge of ball C2 to C3.

    Best Regards,

    Fred

  • 0 in reply to Fred Archer
    TI__Guru** 102885 points

    Hello Fred,

    I will assign someone more familiar with this gauge to make suggestions. I would not use the SLUA903 since this app note is for flash based gauges to complete the learning cycle. ROM gauges require less setup, just confirming with the GPC chem tool to make sure you get the best fit ROM gauge.

    You can use the load select set to 0 (default 1) to try to reduce SOC jumps with a pulsed load, it will use the average discharge from the previous cycle for SOC estimations, instead of the present value.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Fred Archer
    TI__Guru 52996 points

    There are two issues that are intertwined:

    1. The ChemID accuracy. The GPC tool gives a general estimate of max. DOD error. What this means is that assuming the log file was collected correctly (with the required relax phases etc.), the tool calculates that for a given OCV from the log file (which is estimated using the log file so this requires precise voltage and current measurements *at room temperature* (as close to 25deg.C as possible), DOD is off by up to X percentage points. (=abs(100*max(ChemID DOD[OCV] - log file DOD[OCV])) over all OCV[DOD]. With DOD = 0.0 to 1.0.

    2. Load prediction and gauging results for a dynamic load.

    If the DOD error is <3% percentage points then chances are that the gauge can still reasonably estimate cell resistance and capacity and QMax (hence perform its intended purpose). 5% still may work. However, it often depends on where in the DOD range this error occurred and that isn't visible in the GPC results. It may be outside DOD that you use in your system and then it doesn't really matter. Or it may be right where it's critical for your application. The tool doesn't tell this, hence we have a rule of thumb. At a minimum you must make sure that the charging voltage is compatible with the gauge. If the charging voltage is higher than what the gauge's ChemID supports, then DOD will saturate at 0 and the gauge won't be able to perform well (or not at all in some cases).

    I would chance it with the 5% option and make sure that you don't charge higher than what the TRM specifies for this ChemID. Make sure you run a really clean learning cycle before using gauging results. Both QMax (first) and Ra (second) must have been learned before using the gauge. This is critical.

    Also make sure that the gauge is configured for your load. If it is configured to predict a C/5 load and your application suddenly applies 5C, then it doesn't matter, if the ChemID is a match - the gauge will not be able to immediately adjust to wild swings. It has various features to deal with this but that goes way beyond what we can explain on E2E. Check the TRM about load select and make sure that this is configured correctly for your application. Also check the delta voltage in the state class to make sure that the gauge adjusts the prediction termination for a load spike so that it doesn't overestimate capacity for a highly dynamic system.