bqEASY Calibration with the bq20z65

Thanks a lot for your prompt reply Jackie!

If that's the case, I'll just follow the steps outlined in the bqEASY wizard and not perform Pack Voltage calibration.

However, since this post, I'm running into an issue which is preventing me from completing the bqEASY wizard - see here: http://e2e.ti.com/support/power_management/battery_management/f/180/t/85360.aspx

I'm hoping you can help me yet again...

Thanks again.

Hello!

I'm also using the bq20z65 and I try to do calibration.
I followed the SLUA391 for production flow. I can flash the Gaz gauge, it works.
But now for the calibration part I'm not sure of the order of what to do and where to plug the power supply.

In the SLUA379E it's written that the calibration is done with these following steps:

/with the power supply instead of the battery cells

Coulomb Counter Offset
Bit 2 ADC Offset
Bit 3 Temperature, Internal
Bit 4 Temperature, External 1
Bit 5 Temperature, External 2
Bit 6 Current
Bit 7 Voltage

/with the battery celll

board offset calibration.

Is that right to do like this for production calibration ?

Regards,

Alex

hi Alex,

You will need a string of resistors and a power supply to simulate cells at the input pins of the device during the calibration section that requires a power supply. Refer to the user guide of the  device EVM downloadable from our website for more details.

thanks

Onyx

Thanks Onyx,

Indeed, in the guide for the EVM it's written to connect the voltage source with resistors simulating the cells at the 1N, 1P, 2P, 3P (for 3 cells). It's written to connect a current source between the 1N and Pack- to give a current of approximately 2Amp for calibration.

1) But during calibration, is that ok if the current flows trough the sense resistor (2 Amp) and at the same time the voltage source applies a voltage from 1N to 3P ?

2) And why use two MOS (charge & discharge MOS) if they are on the same line ? Because I have to choose a discharge MOS bearing 40Amp (for my application) and then, the charge FET being on the same line, this last one must bear 40 Amp as well and then I have two big MOS instead of one. (of course the charge current is much less than 40Amp, around 4Amp)

Best regards,

Alex

Hi Alexandre,

1. It is ok, but in this case you will have to use the bypass fet option when calibrating to ensure the fets are turned off. That way the only current the sense resistor sees will just be that from the current source.

2. If you are going to have 40A discharge, both FETs will have to be rated for 40A as the current will flow through both. During charge and discharge, both  FETs are turned on. You have two FETs because there are some functionality of the IC that are charge or discharge controlled and in instances like that the device will need to turn on/off either of the respective fets. Please refer to the technical reference manual for such instances.

regards

Onyx

Thanks Onyx,

I have got an other question. In SLUA379E–May 2006–Revised August 2013, it's written:

"iTemperature is in Kelvin/10 units, so the argument is: 10 * (Celsius + 273.15)"

On the line "IError = WriteSMBusInteger(&H62, iTemperature)", do I have to put the argument iTemperature in Kelvin*10 or in kelvin/10 ? or in Celsius ?

Thanks,

Alex

hi Alex,

Calibration within the gauge for temperature is done in units of kelvin/10. So your temperature which is measured in Celsius will have to be converted to kelvin/10 according to the formular by your code.

thanks

Onyx

Hi Onyx,

But to obtain kelvin/10 (which 10 times smaller than the real kelvin value) we must do 0.1*(celsius +273.15) and not 10*(celsius+273.15) ?

Thanks

Celsius+273.13 gives you a value in kelvin. Multiplying x by 10 gives you 10x Kelvin. Since the process uses kelvin /10, the 10 cancels out and the eventual unit is kelvin.

thanks

Onyx

Hi Onyx,

In order to be sure, in the SLUA379E it's written for the board offset calibration (page 6) that "the device under test must be powered from the cells only and no external load or charge current may be applied" but does it implies that we must use the physically cells? Or can we keep the virtual cells done with a power supply and resistances ?

Best,

Alex

Hi Alex,

You can also use the dummy cells i.e power supply and resistor string for calibration.

thanks

Onyx