Bq34z100EVM Problem

Any ideas on why the chip heats up and shuts off when I try to pull power from the battery through the sense resistor? 

Ryan,

I do not see anything wrong in your picture and I do not see any issues on the bench with my EVM setup like yours. You must have a defective device. You can export a senc file and I will try it, but I do not see how the setup could cause the device to get hot.

Tom

Thanks for looking it over for me. 

I have a new EVM  board on the way so lets wait to see if I have the same problem with the new board before doing anything else. 

Do I need to do a Impedance Track learn cycle on my battery? Is there a learn cycle option with this Bz34z100 chip? I don't remember seeing that option in the BQ software. 

Ryan

You do need to run an optimization cycle on the pack, but that would not have caused the device to run hot. You also need to calibrate the device and enable Impedance Track. Refer to pages 29 and 30 of the datasheet.

Tom

So can you give me a laymans quick run down on how I should do a correct Impedance Track learning cycle with this Bq34z100 chip? I read the explanation in the datasheet but it just tells me how to active Impedance Track via the control command.  

Ryan

The process for running an optimization cycle for this device is the same as for our other IT devices. Please checkout this document on the TI website to see it it helps. SLUA597

Tom

Ok I read that document and it did answer all my questions on how to run the test. 

I see I started out the wrong way with a fully charged battery. 

How do I start from scratch with the Impedance Track learning cycle with a discharged battery? 

Yea I deleted that post because I got it working by changing the Terminate voltage back to 2.5 volts. 

At least now I know what was causing the problem now. Thank you. 

Makes sense. 

So should I do this: 

#1. Start out with a fully discharged battery that has rested for 2 hours. 

#2. Issue a reset 41 command.

#3. Issue the IT Enable 21 Command to force the learning cycle again. 

#4. Follow the recommended learn cycle instructions in the SLUA597 guide. 

Sounds easy enough. 

 issue a reset 41 command and an IT Enable 21 command to force the learning cycle to start again with a discharged cell that has rested for a few hours. 

That should work. Note that you are using a LiFePO4 cells, so the rest period may be much longer after charge and discharge. I have seen it take 24 hours, but hopefully yours will be quicker.

Yes indeed it does look like it may take 24 hours for the discharged voltage to become stable. I'm at 17 hours and its still slowly creeping up. 

Is there any reason the CHG & DSG flags should be set at the same time? 

I noticed that the chip on the 2nd evaluation board got up to 75c as soon as I disconnected a load and then reconnected it.

I saw that the CHG & DSG flags were set at the same time and figured that might indicate the reason the chip is heating up internally. Nothing else has been done that could have caused any damage to the board. 

I disconnected the battery from the chip and then reconnected it. It no longer heated up but the current is showing 0 but the CHG flag is still set even though the fuel gauge is in Sleep Mode. Is this normal? 

I did a control 41 reset command a few times but it does not clear the charging flag. 

I'm hooking up a small 2 amp DC Switching boost regulator that takes 3v and converts up to 5v DC. Do you think that maybe this device could be causing a voltage spike when disconnected or initially connected that could be damaging the IC? 

Yes, it is common for the CHG flag to be set. It is defined differently on this device as compared to some of our other devices. It will be set whenever the pack is eligible to be charged.

Any ideas on why the chips on 2 different evaluation boards would be heating up like this? 

If the voltage spiked above 5v when set to 1 cell configuration could that easily cause this issue? 

Only thing I can think if is that this DC switching boost regulator is pushing out a voltage spike higher than 5v when initially connected or disconnected to the battery pack terminals on the evaluation board. 

If this is the case would just replacing the Bq34z100 chip fix the whole board or would a small voltage spike above 5v damage other stuff also? I'm not sure if the voltage is spiking above 5v but its really the only thing that could explain why the chip begins heating up like this. 

I just disconnected the battery from the evaluation board and plugged in the load DC regulator to the battery pack connection on the evaluation board as it should be. Then I plugged the battery back into the evaluation board and this time it did not heat up and its working fine again. Hummm. 

Voltage spike maybe? Any ideas on what could actually be heating up internally pulling .5 amps @3v without completely killing the chip considering it remains operational and transmitting data? 

Let me know what you think. 

I have damaged these EVMs before by applying too much voltage to the BAT pin. In some cases the bq34z100 gets damaged and we have also seen the FET in the REGIN power path damaged. If you have a specific setup and procedure to induce the condition, then I can try it here. Also, please extract you senc file and I will use it.

So if I wanted to fix the first evaluation board I would need to just order a new Bz34z100 chip and what is the part number the FET that might possibly also be damaged? I'll order one of those also. 

I'm not sure how to get you to replicate the buck dc converter I'm using except to send you one of them. Everything else is pretty standard, just a 3.2 LIFePo4 battery and the EVAL board setup for a single cell monitoring. 

On the board that is still working, how could I tell if the FET was damaged? 

Ryan,

The FET is located in the REGIN power path and we used a 2N7002. You could use a similar FET that you may have available. It just forms a simple linear regulator.

Tom

Hello Tom and Ryan,

It seems that I've encountered the same problem :

On BQ34Z100-EVM, designing to work with a LiFePO4 1S battery (1.4Ah), sometimes the BQ34Z100 crash (stop communication with EV2300) + about 600mA drawing from battery + BQ34Z100 temperature increase.

It arrive sometime when switching the 5V laboratory power supply for the charge circuit, and sometime when connecting the load...

As I don't really understand, I was thinking of a BQ34Z100 possibly endamage with first tests, I've replace the BQ34Z100 with a new one on BQ34Z100EVM. But there still the problem.

Do you find a solution? Or the reason why this problem occur?

Regards,

Hubert

Hi Tom,

I've changed the 2 2N7002. But still have the problem.

I've not try to measure pin voltage, but noticed that the problem seems to come when applying the load while the device come from disable state (I'm using a circuit for driving CE high, when the load is connected).

Hi Tom,

I've measured BQ34Z100 pin and measure a peak on SRN, a little more than 300mV.

The pack was supplying a step-up and the step-up had a 50µs peak current of about 3A (so, with my Rsns=100mR, sometime goes under -0.3V on SRN).

When I replace the step-up by a pure resistive load, there's no peak and no problem.

Now, I've put a filter between the pack and the step-up and the problem is resolved.

Thanks,

Hubert