BQ24103 mistery

It so happens that this charger is used with new battery pack that has intelligent battery protection/gas_gauge/monitoring circuit that has MOSFET dis/connect switch that turns on/off when battery is low (old pack only had overcurrent protection). What is the chance that (buck) charger gets blown up because it faces weird (switching) load?

Page 18 in the datasheet says "The bqSWITCHER provides a built-in overvoltage protection to protect the device and other components against damages if the battery voltage gets too high, as when the battery is suddenly removed."

Should be ok then. Anything else would have surprised me. After all, such chips are not supposed to blow if a user is a bit clumsy and the battery insertion is done hot and a bit erratic.

If the components around it are the same, especially things like the output cap, and the layout is the same, it's probably time to contact TI's tech support about this.

As far as I am concerned the chip should survive anything from open to solid block of Cu instead of the battery. Apparently it does not. Actually same charger was used in previous product where we had problems with battery contacts for a while. No failures AFAIK. The difference: "old" battery pack had simple overcurrent circuit. New one has too much electronics built-in.

They are not very helpful (so far). We buy only 10k/year (in a good year)... I should have used LT part.....

More than the usual FETs that cut it off in over- or under-voltage situation? Maybe those electronics warrant a closer look then? If there's anything inductive in there, that could cause problems.

Yep. If I expect to need good support I always pick LTC. I have yet to see any other company with their excellent level of support.

My wild guess, and I haven't a clue how, is that the battery electronics is taking the charger output negative. Is your charger protected against this happening?

First, double check the basics: Your circuit does have the output zener, right? You are not screwing around with the CE or VTSB lines, right? And the output from the brick, under load, is solid & clean, right? Any "wrong" above opens you up to potential problems.

Ed

e"

red

I noticed today that under certain circumstances (battery pack going in and out of undervoltage mode - charge/discharge cycling) I can get nice 1kHz sinewave between 1V to 3V. This is with different (purchased) charger, though. I have not been able to reproduce this with our (built-in) charger. Neither was I able to kill our charger. When I was playing around with the board, I received a present from production - board with another blown charger. Grrrrrrrrrrrrrrrrrrr. To answer ehsjr questions:

Your circuit does have the output zener, right? - yes and no... According to TI the chip with letter A does need a zener. We have been using it without zener until now. The zeners were "reworked in". I think we have lost one reworked board. You are not screwing around with the CE or VTSB lines, right? - what is it? this is industrial instrument, powered either by the battery or by 12V (brick) And the output from the brick, under load, is solid & clean, right? - drawing 3-5A from 7A brick made by... I forgot, some reputable company, should be solid and clean enough...

I am pretty sure that battery pack "intelligence" messes up our charger.

Recommend you start looking these things BEFORE production gets them to go brzzt.

Much easier to diagnose and trigger to waveforms during the event, than a few days later.

RL