BQ2031 falls into unpredictable state

Please monitor the waveform of charge current and LED signal. If we see some pulse, the pre-condition is there but not stable due to low battery voltage.

We can also monitor the battery voltage to see if the battery is still being slowly charged up.

I'm facing the same issue using the DV2031S2 evaluation board.

Both LED2 and LED1 pulses at 840ms interval in a non-documented way. Is there anything to do with Loop stabilization as mentioned in slua019.pdf?

Since it's an evaluation board I believe current and voltage loop stability was already calculated and implemented correctly.

This issue made me loose all the confidence in the solution. Have anyone experienced the same issue?

Regards,

Just giving more data...:

When a battery to be recharged is connected and input power is applied (24V), BQ2031 enters in an unpredictable state with both LED1 and LED2 blinking in aprox. 840ms interval.

Every-time LED pulses, the MOD (PWM) output will pulse. The current sense resistance RSNS (R20) shows triangular current pulses waveform with 5~7A of peak. The BQ2031 will start correctly sometimes when I reduce the input voltage down to 9-10V and wait for a few unpredictable LED pulses/cycles.

The device will stay in the described scenario for an unknown amount of time using input voltage of 24V.

I also tried to limit the inrush current using the current regulation from my power supply but it did not make any difference.

The demo board is working normally with 4 other batteries. Is it a current or voltage loop stability problem?

Have anybody had any similar issue with this device?

Appreciated.

Hi Alex

I have a similar problem when I connect certain batteries to the DV2031S2 eval board.  After powering up the board I connect a Battery with an open loop voltage of about 10.8V (close to deep discharged). What happens then is, that all leds turn off and I can hear a faint periodic clicking noise. If I use an oscilloscope, i can see that all leds are beeing activated for a very short period every (make an educated guess...) 840ms! This goes on forever.

Because this doesn't happen with batteries that are in a good condition (~ >11.5V open loop voltage) I assume that the BQ2031 seems to have a problem to pre-qualify deep discharged batteries. If I take such a "bad" battery and pre-charge it with a power supply in CC-Mode (500mA for about 5-10 mins), it also works with the BQ2031 afterwards! Maybe theres a chance to tweak some resistor values or jumper settings in order to solve this problem, but I don't know it by now.

Cheers...

Really appreciated you sharing experience on this.

Unfortunately I had no time left to keep investigating/studding it, and since TI believes it's not a tech support question, I had to move switch over to a different solution. 

One of the reasons why I had chosen this device is the fact that it's been around since 1999.

My wild guess is that it has to do with loop or voltage compensation (VCOMP and ICOMP pins). The formula for one of the compensation loops takes into account input Voltage and battery' internal resistance and capacitance. I'm able to make the device start pre-qualify correctly if  the input voltage is reduced bellow 12V, in other words, I'm variating the VIN parameter of the equation. I just want batteries to be recharged, that's what I expect form a recharger :-). I don't care about any possible value for battery's internal parameters (unless I would like to know charge state using impedance but it's not the BQ2031's proposal). My feeling is that I may be calculating compensations over and over and still going to find a battery that will put the device into a unpredictable state.

Batteries near deep discharge are the ones that I would fully rely on BQ2031.

Regards.

Looks like you're right and its something with the loop compensation. I just changed the given compensation network on the DV2031S2 (100nF, 100nF/0Ohm) to the one mentioned in slua019.pdf, Page 10. It was a lucky-shot, but I thought it's worth trying! Even though it's not working completely stable, it's at least starting up with almost any given deep discharged battery I can find around here. I will try to recalc the values and stabilize the charging process.

What other solution did you find that worked for you?

It's not an of-the-shelf solution. I decided to build my own recharger using the microcontroller already in the project (aggregate more functionality to it) to control the PWM. I'm using DS2031S2 and www.st.com/stonline/books/pdf/docs/12476.pd as a starting point. Basically a SMPS with Buck topology...