How to charge the super capacitor very fast?

Hi guys
I have to charge 400F super capacitor up to 4,5V in 180 sec.
What should i use?
How about bq24640?
Is there anybody who used bq24640?
If there is, how much are bq24640 and transistors heating?

Charging time is normally regarded as five RC time constants.

You want a charging time of 180 sec which gives a time constant of 180/ 5, or 36.

The resistance in the charging loop can be 36/ 400, or .09 ohms.

.09 ohm is the maximum you can have in your:
power source
+ switching device
+ wiring
+ capacitor.

If your power source is 4.5 V, then the charging current will start at 50 amps. It will drop below 15 A after 36 seconds (one TC).

Your cables should be 6 or 8 gauge copper to handle the current.
Can your capacitor leads handle that?

If your switching device is .09 ohm then it will temporarily have to dissipate I x I x R, or 50 x 50 x .09, or 225 watts. For the first 10 seconds, say.

If your switching device is .045 ohm then it will dissipate 50 x 50 x .045, or 112.5 W.

Mosfets are available that can handle this, with proper heat sinking. You may need more than one. Its leads must carry 50 amps for a few seconds.

Are you making one of those railguns?

I am just curious about how you calculate " start at 50 amps"?
Thank you so much.

First remark, I really wonder, why it's so difficult to place a question in the correct forum?

Secondly, constant current would be the obvious method to charge a supercap effectively. Charging it through a resistor implies to lose half of the input energy, only a (ideally lossless) switching converter can charge it from a voltage source without huge losses. If current controlled, it would source 10 A. Alternatively, a constant current/constant power characteristic may be appropriate.

Thirdly, constant current is in fact the standard operation mode of the bq24640. Because it's just a controller without power transistors, most properties depend on the detailed circuit designs. You would want to refer to the TI datasheet and reference design to answer your respective questions. 10A charging current sounds feasible, however.

Sorry, I was basing my figures on a power source is 4.5 V. You probably have a power source at greater V.

You can charge the cap from a 12 V source through a slightly greater resistance. Your charging current will start much lower than 50 A and the current will drop more gradually.

Or from a 24 V source through an even greater resistance.

There's a formula to determine what current needs to go into the capacitor and for how long.

I was using easy theoretical calculations where charging ends automatically when the cap rises to the source voltage. At that point charging current is zero.

Considering the purpose of a supercap is energy storage, you won't want to charge it through a resistor. The efficiency reduces to Vcap/(2*Vb).

Yes. We don't need any more resistance in the charging loop than there already is.

I hadn't looked at the datasheet for bq24640 until now. It appears ideal for the purpose. 10 amps for 180 seconds is a much better charge algorithm.

I'm behind the times, I'm afraid.

I still remember the novelty of seeing 1 F capacitors available in electronics mail order catalogs.

You could buy the chip, or....

build the circuit yourself. all you need is: 1.) voltage reference 2.) an opamp 3.) current sense resistor 4.) mosfet

I did something similar to this a few years ago. except I did 15A constant current.

what type/name is your super cap?
What serial resistance have this super cap?

---------- Post added at 05:05 ---------- Previous post was at 05:04 ----------

what type/name is your super cap?
What serial resistance have this super cap?

hi
thanks for reply
I have already bought chip and other components.
I'm curious about your project.
Please share your previous project