Battery/Supercapacitor LED flashlamp?

[cupoftea]

Hi
We are doing a high power LED flashlamp, 5 seconds ON, 5 seconds OFF. Power is drawn from a 2S Li Battery (7V4). So the current drawn from the battery is 18 Amps for 5 seconds, then nothing for 5 seconds, and so on, etc , etc.

We wish to introduce 2 series 50F , 3.1V supercapacitors (6v2) to smooth out the current draw from the battery.

So do you agree, what we must do is first make the battery charge the supercap up to 6.2V.
Then split the flashlamp into two equal flashlamps.

….Then we simply draw a constant 9A from the battery….for 5 seconds we send the 9A to the one flashlamp from the battery....and the same power to the 2nd flashlamp but from the supercap...then for 5 seconds we charge the supercap with 9A.....?...etc...to be continued...

So basically, we draw constant 9A from the battery.....for 5 secs send it to flashlamp1...then for 5 seconds send 9A to the supercap...........the supercap will augment the light by giving 9A to the flashlamp2...which has the right voltage so that it gives the same power as the flashlamp1, when it flashes.

Does this sound the best way?

 

[dick_freebird]

5 seconds on is not a "flash". You're well past thermal
time and into stabilized operation.

People do this kind of thing, using the supercap to
cover up the slow roll-on of battery current to pulsed
load.

But I don't think they're expecting, or getting, 5 seconds
of current (nor should they need it).

You ought to know what the battery transient performance
looks like, fresh and aged, full and low charge, to scale
any supercapacitor volume. And you might then look at
-that- which can still be more leisurely than a simple cap
and maybe crutch the crutch.

As a primary charge delivery element, I think any available
supercap will come up short. 1F and 9A, if you accept 10%
voltage droop as the discharge tax, only gets you dt=CdV/I
or =1*0.31/9 = 34ms of "coverage", after which time you'd
better be taking it from the battery (meanwhile, it's already
seeing the cap want some too).

That's nohow close to 5 seconds.

I'd be pushing on that "5 second flashlamp" story, because
it's either not one or it's not the other, and the hang time
drives a lot of choices and/or cost.

 

[wwfeldman]

i took a quick look at one 50F 3V supercap. maximum ESR is about 13 mohm
(i picked the one with the lowest ESR)
RC time constant (for two in series - 25 F and 26 mohm) is 0.65 seconds.
full charge in 5 time constants is 13 is 3.25 seconds

if you pull 9 A out of a 25F cap for 5 seconds dv = I/C dt or 9A*5s/25F = 1.8 V (much bigger that dick_freebirds reasonable 10%)
assuming a constant current

using dick_freebird's analysis:
dt = Cdv/I, dt = 25F 0.6V/9 = 1.7 seconds (assuming 2x50F 3V caps in series)
still not goint to make 5 seconds on

as i read what you wrote:

….Then we simply draw a constant 9A from the battery….for 5 seconds we send the 9A to the one flashlamp from the battery....and the same power to the 2nd flashlamp but from the supercap...then for 5 seconds we charge the supercap with 9A.....?...etc...to be continued...

So basically, we draw constant 9A from the battery.....for 5 secs send it to flashlamp1...then for 5 seconds send 9A to the supercap...........the supercap will augment the light by giving 9A to the flashlamp2...which has the right voltage so that it gives the same power as the flashlamp1, when it flashes.

Does this sound the best way?

charge supercaps
loop start:
power flashlamp 1 from battery, and power flashlamp 2 from supercap for 5 seconds
recharge cap from battery for 5 sec with both flashlamps off
repeat loop

plus losses from switching current path out of batter and to either a flashlamp or capacitor

probably going to draw closer to 10A from the battery
with glitches every time you switch the current

 

[cupoftea]

Its a New thread but on supercaps so please may i ask this?.........thought i should put it here otherwise its like double posting(?).
________________________________________________________________________________________________________
Hi,
We wish to pulse a laser diode block with 20A for 250ms every 2 seconds.

(LTspice and jpeg as attached)

We cannot draw this pulsation from the 2S Lithium Battery (26650's) that we have , so we will charge up supercaps with the battery instead, and draw the pulses from the supercaps instead.

We will use the VinaTech 25 Farad, 3V, 13mR ESR supercap…2 in series.

VinaTech 25 Farad, 3V, 13mR ESR supercap
https://www.vinatech.com/winko/data....3_85%B5%B5_%C1%B6%B0%C7%28R0%2C210201%29.pdf

Do you think the supercap can handle these pulses?…(datasheet doesn’t allure to pulse rating) its only some 0.7W of power in each supercap

 

[FvM]

High drain Li batteries can supply 40 A and more. Why supercap?

 

[dick_freebird]

Batteries can have leisurely roll-on of current depending on
the timescale of interest. There are outfits making combo
cell+supercap assemblies for high dI/dt pulsed loads.

I see the "flash" duration has become more reasonable
as this discussion has progressed. But the supercap is
all about the leading edge.

I imagine that before this is all done, somebody will want
to specify and measure "optical waveshape", which is what
the fussing over caps, supercaps and battery is about in the
end. If the LED bank is white LEDs you might find that you
have very little control over the optical trailing edge, as the
phosphors' "hang time" can be of the same order as your
present pulse width number. And that being the case, maybe
you are not too concerned with the rising edge either. What's
the "quantity of interest" here anyway? Just optical Joules
delivered? Some sort of "illumination fidelity" meanwhile?
Or can you just charge a fat-ass cap to a Coulomb value
and let it rip, getting the net energy and not caring about
shape so much?

I notice a bad habit of accepting "requirements" at face value,
only to have them change later. You might save wasted effort
by challenging and analyzing early, and making sure they
are both real and agreed upon. If you have the perspective
and basis to challenge.

 

[cupoftea]

High drain Li batteries can supply 40 A and more. Why supercap?

Thanks, because we will also somtimes be using high capacity cells , which cant manage the high current pulses.