BQ25758A: Full current output vs CE low duration

1. A bit detailed about the task: there are really high current pulses - 40 - 300A, 50 - 200 us, with 1% (max) duty cycle. Caps voltage (charged) is 36V, worst-case min caps voltage is 27V (except initial charge - surely initial charge starts from 0V). Calculation says, that min chatge current is 3A, 5A - defined (seems will be enough).

2. Looking onto pulses data, I thought, that to hold charger for a while (when such strong load pulse is present) is good idea and blessed solution (hidh dI/dt transitions - interference, noise, unpredictable behaviour etc etc). But if the IC will just switch to CC mode and nothing more - it's fine. A question - what "will be better" if I'll leave CE low while pulse is active? And one more - one word in your reply is disturbing me a bit - "probably". Um?

3. What about the IC's "sisterships" (bq25756 for example) - if I'll pull CE low "forever" (and enable charge in I2C regs - surely) - will bq25756 "immediately" start charging process? In worst case it will be 5 ms (50us-pulse, 1%-DC -> 5ms period, 200Hz freq) for all charge process. Or has bq25756 (-750, -820) different internal logic (batteries charge processes are not so fast as in my case :) ) and there is long duration between drop of voltage on FB pin and Full Charge Current state appear (full current applied to output)?

4. And one more question - I2C control and harware control - what to choose? Hardware looks more reliable from my "harware designer" point of view :), but may be I2C is more preferred - from your (IC designer) point of view?

And one more question - could you tell me what are key differences between bq25758 and bq25758A? Surely I can read three datasheets "letter-by-letter", but.. if possible - a small summary. Please :)

Hello Maxim,

1. Thanks for the new information. The BQ25758 will be fine during this time.

2.

Maxim Shtyrov said:

A question - what "will be better" if I'll leave CE low while pulse is active? And one more - one word in your reply is disturbing me a bit - "probably". Um?

Thanks for being patient here and let me rephrase this. I think it will be better to leave the charger charging all the time. It will go into CC mode once the capacitor voltage starts dropping and the current exceeds the IOUT current value.

3.

Maxim Shtyrov said:

What about the IC's "sisterships" (bq25756 for example) - if I'll pull CE low "forever" (and enable charge in I2C regs - surely) - will bq25756 "immediately" start charging process? In worst case it will be 5 ms (50us-pulse, 1%-DC -> 5ms period, 200Hz freq) for all charge process

I'm not quite sure what you mean by this. Can you rephrase this question? The BQ25756 will have the same time to start charging as the BQ25758.

Maxim Shtyrov said:

Or has bq25756 (-750, -820) different internal logic (batteries charge processes are not so fast as in my case :) ) and there is long duration between drop of voltage on FB pin and Full Charge Current state appear (full current applied to output)?

The battery chargers will terminate charging once the battery/capacitors have been fully charged. Once the battery/capacitor voltage has fallen below the recharge threshold, the charger will start charging again. Although, the bit EN_TERM=0 can be set to disable termination.

4. Either I2C or hardware will work just as well here. It will depend on what resources your system has available. I think using the BQ25758A with I2C would be the easiest solution here.

Maxim Shtyrov said:

And one more question - could you tell me what are key differences between bq25758 and bq25758A? Surely I can read three datasheets "letter-by-letter", but.. if possible - a small summary

The BQ25758A has a better transient response for USB applications, but both chargers will work equally well in your case.

Let me know if this answers your questions and let me know if you have any other questions as well.

Best Regards,
Ethan Galloway

Hello, Ethan.

Thanks a lot for detailed reply.

About bq25756 charger and bq25758 converter - let's have a look into datasheets and I'll try to explain what I mean.

1. bq25756 charger:

1.1. Section 7.6 Timing requirements, p.15.
"Battery Charger" block - "Deglitch time for recharge threshold, VFB falling" = 200ms.
Did I understand properly, that this is the delay between battery voltage falls below recharge threshold and the restart of charge process (CE is low "forever")? 
And if I'll disable charge termination, what will be? IC will stabilize output voltage (CV mode), and will be in this state as long as I wish (think I should switch safety timers off, yes?), then, when VFB will fall below recharge threshold - what will be? 200ms delay? Or will bq25756 work like bq25758 (CV->CC transition  - see 2.2)?

1.2. Section 9.2.1.3 Application curves, p.68.
Figure 9-7 "Charge Enable via CE Pin with 5-A ICHG".
I can see, that ICHG appears ~150ms after CE pulled low.

So, let's summarize this part - if bq25756 behaviour with disabled charge termination will be like bq25758, that's good, bq25756 could be used - let it be "reserve trench" for me.

2. bq25758 converter:

2.1. Section 6.6 Timing requirements, p.13. 
There are no data about delays between VO_SNS (VFB) fall and IC "active" - yes, that's ok (it's DC-DC converter, active continuously).
But let me return to CE pin logic - what about the delay between "CE pin low" and "IC is active" events. Looking into bq25756 datasheet (see above), I have some doubts, while bq25758 is "sistership" of bq25756. It's the first question.

The second one:

2.2. Section 8.2.1.3 Application Curves, p.55.
Figure 8-10. "Output Short Circuit Response".
Stop, stop, stop - "CC mode" or "Short Circuit Response"?? What "short circuit" is meant? A-a-ah!! This figure is about switching from CV to CC mode, right? And CV->CC transition will take... (my God, you have to kill for such pictures... 800% scale and you still have to put a ruler to the monitor to estimate the time... Damn it...) ~700us and I_load will be interrupted for ~400us. Am I right? And how will the IC jump back - CC to CV?

Part summary - if "IC active" vs CE->low delay significantly differs from bq25756 (could you please determine at least an order -ns, -us, - ms...) - very good, this way is possible. If CC->CV transition at least not longer than CV->CC - very good, this way also open for me.

Hello Maxim,

1.1

Maxim Shtyrov said:

Did I understand properly, that this is the delay between battery voltage falls below recharge threshold and the restart of charge process (CE is low "forever")? 

Yes, this is correct.

Maxim Shtyrov said:

And if I'll disable charge termination, what will be? IC will stabilize output voltage (CV mode), and will be in this state as long as I wish (think I should switch safety timers off, yes?), then, when VFB will fall below recharge threshold - what will be? 200ms delay? Or will bq25756 work like bq25758 (CV->CC transition  - see 2.2)?

If you disable charge termination, the BQ25756 will never stop charging. The deglitch time won't apply here.

1.2

Maxim Shtyrov said:

So, let's summarize this part - if bq25756 behaviour with disabled charge termination will be like bq25758, that's good, bq25756 could be used - let it be "reserve trench" for me.

Yes, the BQ25756 will look like the BQ25758 during operation.

2.2. I don't think Figure 8-10 will apply to your application. Figure 8-10 shows VOUT being forced to fall all the way to 0V. This won't happen in your case because you have a large capacitance on VOUT.

Maxim Shtyrov said:

if "IC active" vs CE->low delay significantly differs from bq25756 (could you please determine at least an order -ns, -us, - ms...) - very good, this way is possible. If CC->CV transition at least not longer than CV->CC - very good, this way also open for me.

Can you clarify the question here? "CE->Low" to "IC Charging" will be about the same for both parts. But, the charger transition from CC to CV mode and CV to CC mode will be significantly faster.

Best Regards,
Ethan Galloway

I'm sorry Ethan, need much time? Or just forget about my replY?

Hello Maxim,

I think there may be an issue with the communication here. Your last message that I saw was on Jan 11th. My last reply was Jan 13th. Did you get my response or have you sent a new response since then?

Best Regards,
Ethan Galloway

Hello, Ethan.

Thank you. Seems there is only one question left.

Ethan Galloway said:

Maxim Shtyrov said:

if "IC active" vs CE->low delay significantly differs from bq25756 (could you please determine at least an order -ns, -us, - ms...) - very good, this way is possible. If CC->CV transition at least not longer than CV->CC - very good, this way also open for me.

Can you clarify the question here? "CE->Low" to "IC Charging" will be about the same for both parts. But, the charger transition from CC to CV mode and CV to CC mode will be significantly faster.

What I mean. Bq25756 charger. Delay between "CE-low" sygnal and "Charge" state (output current appears) is shown in datasheet and it is ~180ms (p.68). Bq25758 is not a charger, just a DC-DC converter and it should not check a number of charge parameters, it should just start switching, I thought, that for bq25758 delay between "CE-low" and "current on output appears" events could be significantly less than for bq25756. 
In another case (if inner logic unit is the same for both chips) meant delay will be approximately the same - as for bq25756 as for bq25758. So, what case is actual?
It's the first.

The second. Could you clear: 

Ethan Galloway said:

But, the charger transition from CC to CV mode and CV to CC mode will be significantly faster.

in this phrase "charger" means both chips - 25758 and 25756, yes? Or with word "charger" did you mean bq25756 only? I'm confused a bit.

And one more question about bq25756.
To make it work like bq25758 I need to shut down precharge, charge termination and stop charge timer: 
• EN_PRECHG = 0
• EN_TERM = 0
• EN_CHG_TMR = 0
Ok. Clear.
But in section "8.3.4.6 CV Timer" I've found, that I need to disable CV timer also:
CV_TMR = 0.
Will this be enough to "turn" bq25756 to bq25758? Or did I miss somethig else?

Best Regards.
Maxim Shtyrov.

Hello Maxim,

Thanks for being patient with this. I'm swamped with work at the moment. I'll get back to your question next week.

Best Regards,
Ethan Galloway

Hello Maxim,

Once again thanks for being patient with this. Monday was a US holiday.

Maxim Shtyrov said:

What I mean. Bq25756 charger. Delay between "CE-low" sygnal and "Charge" state (output current appears) is shown in datasheet and it is ~180ms (p.68). Bq25758 is not a charger, just a DC-DC converter and it should not check a number of charge parameters, it should just start switching, I thought, that for bq25758 delay between "CE-low" and "current on output appears" events could be significantly less than for bq25756. 
In another case (if inner logic unit is the same for both chips) meant delay will be approximately the same - as for bq25756 as for bq25758. So, what case is actual?

Once again, thanks for being patient with this and I should've clarified this better. I was referring to both the BQ25756 and the BQ25758 when I said "charger". I'll try to start referring to the both of the ICs as converters to remove the potential confusion.

Maxim Shtyrov said:

in this phrase "charger" means both chips - 25758 and 25756, yes? Or with word "charger" did you mean bq25756 only? I'm confused a bit.

Yes, "charger" means both chips.

Maxim Shtyrov said:

To make it work like bq25758 I need to shut down precharge, charge termination and stop charge timer: 
• EN_PRECHG = 0
• EN_TERM = 0
• EN_CHG_TMR = 0
Ok. Clear.

Yes, this will be enough to make the BQ25756 act as a constant voltage source like the BQ25758.

Maxim Shtyrov said:

But in section "8.3.4.6 CV Timer" I've found, that I need to disable CV timer also:
CV_TMR = 0.
Will this be enough to "turn" bq25756 to bq25758? Or did I miss somethig else?

Thanks for noticing this. The CV_TMR isn't for the BQ25756. It's for the BQ25751 (a lead acid battery charger). I'll remove this from the datasheet.

Let me know if you have any other questions.

Best Regards,
Ethan Galloway