BQ34Z100-G1: Internal impedance value

Matt, sorry, I looked at the BQ34Z110 not the BQ34110. Thank you for info, I will check it out.

So just to summarize, would it be possible to estimate the current impedance somehow during the product life, without having to disconnect the circuitry? this is only for display purposes.

Hello Matt,

I have one more question regarding the BQ34110 chip. It does say in the data sheet that it is dedicated for UPS backup systems. The thing is that it can accept only 32 A currents. It does say that scaling is possible, but never explained. Also, even 32 A for a shunt resistor is a lot to take. Could you elaborate a bit more?

Let me add why I am asking: the application under consideration is the one in which a lead acid battery is connected of capacity around 200 AH. The battery serves as a UPS backup but its discharge rate can be even 500 A (the whole battery can discharge within 10 minutes). Is this chip suitable for such application? I cannot imagine using shunt resistor here (although its possible) and was wondering about connecting a galvanically separated sensor to the BQ34110.

Hello Matt, thanks for answer.

Even assuming I physically managed to put appropriate sensing method into the system (shunt resistor, current transducer or a hall current sensor IC), the current resolution for the BQ chip (14 bit ADC), for +-500A would be around +-250 mA, which I find very low at situations where the UPS system is being kept charged (99% of the time). Is there any way to force the BQ34110 to operate without measuring current, but only by using rails voltage and learning load?

Hello Matt, thank you for answer.

To summarize, could you please answer those questions?

1. Can the EOS functionality be used without the SoC functionality, or do they have to be used together?
2. Can the EOS functionality be used when there is no shunt resistor (or other components used for current sensing) in the circuit, but only the learning load resistor?
3. Using the EOS functionality, is it possible to obtain the impedance/ resistance value of the connected cells?

Like you said, SoC is not very useful, I agree. It is a good addition, but the current impedance is the most important in this application. The system will be neraly always charged and in rare situations fully discharged in around 10 - 15 minutes. The current will be so high, that I would really like to avoid even trying to measure it, as it would introduce additional costs and/or power losses.

I would appreciate your further help.

Thank you for answer Matt,
This cleared thing up. This will be a tough one... I gues a shunt resistor on the RLEN will not do for the internal impedance track algorithm?

Hello Matt,

I have been thinking about this application a bit further. Since as you said, the SoC mechanism is not needed for the EoS mechanism, thus, they can work separately, but the Current() function is still needed, I have came up with the following idea:

Do you think such changed circuit would work correctly for EoS only? I have removed the shunt resistor from the cells series connection completely, and added it to the discharge load circuit. This way the current will be possible to be measured during controlled discharge. Of course I understand the drawbacks, but just for now, lets assume that during that discharge there will be no other discharge patches available. Would this work as a workaround? Or is the Current() function needed during charging as well?

Otherwise, I am out of ideas. In this application there will be multiple 12 V batteries connected in series (total voltage of around 400 V) with a common charger for all of the series batteries. Each battery would have one BQ34110 controller attached. Measuring the total current of around 500 A through a shunt resistor (or other sensor) is not really an option here, it will be too expensive..

I would appreciate your further help.

Hello Matt,

I have some new information regarding the design. The hack with the shunt resistor in the load circuit might not be necessary after all, since there is a Hall current sensor available for the whole series pack:

For this application, the sensor would output around 4V when 500A current was detected and -4V for -500A accordingly. Like you mentioned, the input range for the BQ device is +-125mV. In that case a voltage divider would have to be added in order to divide the voltage by 32 (4 / 32 = 0.125). I would then connect the hall sensor output through voltage divider to the SRP pin and ground the SRN pin. Is this approach correct?

Assuming 14 bit ADC in the BQ34110 device, this should allow to measure current with around 30.05 mA (or 60.1 mA?). Is this enough for the SoC and EOS algorithms? Would appreciate your feedback.

Hi Matt, thank you for answer.

I find it a bit confusing about the SRP and SRN connection, since normally the local BQ34110 GND i tied to SRN pin. Is this circuit correct then? Am I missing something?

Hi Matt,
Ok this is more clear now. I have already seen the video, you linked it before, thanks. Ok, so the last question point is that with the hall sensor, I will not be able to utilize the EOS algorithm, but only SoC. This is due to the fact that there is only one Hall sensor per multiple series connected packs. Each pack has BQ34110 attached. The following block diagram shows the problem:

In this case an analog switch circuit would be needed to shift between the hall sensor for SoC and shunt resistor attached in series with load for load learning... Can this work?

Hello Matt Sunna 

So I have some early version of the BQ34110 circuit:

I would like to know your opinion about it. I have tried to introduce all of the solutions we have discussed about. So, there would be 2 mounting options for this circuit:

1. SoC measurement. EOS is not measured.
2. EOS measurement. SoC is not measured.

In the design there are 10 series packs. Each pack is 60V max, so the whole string is 600 V. Each module (pack) has an EOS measuring BQ34110 and for the whole string there is 1 SoC measuring BQ34110.

I would appreciate your thoughts on this solution.