SUPER CAPACITOR STEP UP POWER SUPPLY

stant     24 vdc , high current ?

It's a bit tricky to get much current out. A transformer-based inverter can work a whole lot better.

What is "high current"? Your input current will be 10 times as high as that, so consider if this is at all practical.

You would need a "boost converter". Some of the products marketed as "energy harvesting" work down to very low input voltages. You might be able to find something that works in one stage, otherwise two cascaded boost converters.

Even 10s of farads at 2.7 volts won't last very long making 24 volts at "high current." Have you done the math on that?

How high a current did you have in mind? For how long?

With some sort of step-up switching regulator, either a boost converter, a flyback converter, or a transformer converter operating in buck.

If you start with 2.7V and if super capacitors have the same voltage vs. charge behavior as regular capacitors, you'll have extracted 90% of the available energy by the time you're down to 0.9V -- that's a considerable range, but at least it's not "almost zero".

Second, what is keeping you from running a number of super capacitors in series? With ordinary capacitors that would be a very tidy way of increasing the available voltage, without changing the available energy per capacitor. You would have to worry about things like equalization, but it might make your circuit easier to design if you start with a series capacitor bank that can store nearly 24V.

Maybe 10 mega-amps for 10 femtoseconds?

"high

How bad is the ESR of those caps?

constant 24 vdc , high currnet ?

"high

some are pretty darn low.

I just got some maxwell 2 or 300F 2.7 volts caps, which are about the size of a D-cell and called "D Cells" or something silly like that.

I shorted one with a multimeter lead and it melted the tip almost instantly.

Interestingly, these caps sound like they're full of liquid, but are extremely light.

there's no question they're dangerous and have an extremely low ESR. It's quite the contrast from the standard memory caps which do nothing if you short them out.

If the capacitor would be loaded with a constant current load, the voltage drop would be linear.

However, if you need a constant current on the 24 V side (i.e. a constant power load), the current drawn from the capacitor must increase as the capacitor voltage goes down. This will further accelerate the voltage drop rate and thus, there is not much point in trying to use close to zero capacitor voltages.

Putting capacitors in series works a lot better with tight-tolerance capacitors.

Typical electrolytics with tolerances in the range +80%/-20% can't be relied on to divide the series voltage equally. Three +80% tolerance parts in series with one -20% tolerance part would put 43% of the total voltage drop across the -20% part.