Charging a supercapacitor

[SoftwareSamurai]

I've decided to start a little side project that involves using a supercapacitor instead of a rechargable battery. While I have read up on using supercaps, I'd just like to confirm a few concepts before I start. (For my experiments, I'm using a 400F 2.7V supercap.)

Q1: Am I correct that I want a constant-current circuit to charge a supercap?

Q2: Am I correct that the charging voltage should start out at whatever the supercap's current voltage is, and then be dynamically raised high enough to maintain the constant-current feed as long as the voltage across the supercap never goes above the supercap's rated value?

Q3: Am I correct that when using a supercap as an energy source, I should always add a little resistor in series (or perhaps a current-limiter) to prevent the load from pulling too many mA?

Any advice is appreciated!

[c4757p]

Datasheet? It differs from one to the next. Some don't require any limiting at all for discharge, as they can't deliver enough current to damage themselves, and only require limiting for charge if you need to avoid the inrush. Even if it does require limiting, I think that a CC circuit is probably gilding the lily a bit for most implementations. A resistor ought to be fine.

[SoftwareSamurai]

The supercap I'm working with is the IOXUS RSC2R7407SR.

[BravoV]

Though just an EE wannabe, take my advise with a grain of salt ..

Q1: Am I correct that I want a constant-current circuit to charge a supercap?

The need for constant current is just merely to avoid overloading the supply part, cause without a controlled charging condition (constant current), the supercap to the power supply is like a short circuit that will fry the charging circuit.

Q2: Am I correct that the charging voltage should start out at whatever the supercap's current voltage is, and then be dynamically raised high enough to maintain the constant-current feed as long as the voltage across the supercap never goes above the supercap's rated value?

If the supply part voltage is "lower" than the supercap voltage, then it will discharge it's energy to the supply circuit, assuming there is no diode to block the backward current.

Q3: Am I correct that when using a supercap as an energy source, I should always add a little resistor in series (or perhaps a current-limiter) to prevent the load from pulling too many mA?

Supercap capable of sourcing huge current, the purpose of limiting the current is just merely a controlled energy discharge, like using a resistor as you mentioned or using a regulator circuit for more precise discharging mechanism.

[c4757p]

Jesus, that's one fancy datasheet... 

With that kind of short circuit current I'd definitely be using a current limiting resistor, or at least a polyswitch. You could start a car with a few of these... Depending on how long of a charge time is acceptable to you, I'd either use a series power resistor, or a CC circuit delivering a relatively high current for faster charge. A 2A charge current will charge in 9 minutes, and a 0.27 Ohm resistor will give you roughly the same charge time.

With such a high charge current and power dissipation, you might also consider a buck converter. It's pretty easy to rig an MC34063 to give constant current. Assuming you're charging it from a 2.7V rail, you'll be dissipating over 3W over the first 2 minutes.

[peter.mitchell]

I'd re-wind a MOT, rectify it, put the rectifier on a big-ass heatsink, charge the cap with that

[Nickk2057]

haha... you guys are nuts when it comes to those supercaps  but i guess those things are great to use (or abuse) in the work environment

[c4757p]

haha... you guys are nuts when it comes to those supercaps

How so?

[smashedProton]

Just make a charge pump ..

[KD0CAC John]

I need to go back to the tool storage , and look for something I bought from Snap On tools while I was a mechanic , and open it up again .
It was a small light jump-start box and used caps for jumpstarting , 12vdc up to 1600 amps , in order to use , it had to be connected to a supplied wal-wart to keep it charged , so not good for keeping in trunk , unless it was connected to the car's charging system , but again that wouldn't work , because if it was parked and needed , the unit wouldn't be charged , so shop use only .

[c4757p]

Just make a charge pump ..

I don't see how a charge pump is going to help you efficiently charge a 400F capacitor. 

[AlfBaz]

I need to go back to the tool storage , and look for something I bought from Snap On tools while I was a mechanic , and open it up again .
It was a small light jump-start box and used caps for jumpstarting , 12vdc up to 1600 amps , in order to use , it had to be connected to a supplied wal-wart to keep it charged , so not good for keeping in trunk , unless it was connected to the car's charging system , but again that wouldn't work , because if it was parked and needed , the unit wouldn't be charged , so shop use only .

If the cap leakage is that bad you'd want to avoid having it in the boot, electron stains are really hard to get out of the carpet 

[Nickk2057]

what I mean is that is one of those short circuits it will blow up in your face... and or cause a fire.... and like KD0CAC John said... "shop use only"... which sounds like a good warning since it will have a good fire extinguisher nearby..

[c4757p]

Eh, just stick a fuse on it. I'd go for a polyswitch - it'll have enough internal resistance to drop the short circuit current even before it trips. Not a big deal unless you go messing around before the fuse.

[Ferroto]

Be careful supercapacitors can deliver alot of current at relatively low voltages, here is a video showing how.




Useful calculator.
http://www.circuits.dk/calculator_capacitor_discharge.htm

[Nickk2057]

Be careful supercapacitors can deliver alot of current at relatively low voltages, here is a video showing how.




Useful calculator.
http://www.circuits.dk/calculator_capacitor_discharge.htm

now see... that is what I was talking about.... just so much energy in those things.... a lot of sparks from just a capacitor (or capacitors)

[SoftwareSamurai]

I decided to try to charge up my 400F supercap and see what happens when I disconnect it and leave it alone for a while.
To charge it, I used an old PC power supply and the 3.3V rail and a tiny current limiter (300mA). I connected my meter across the cap and watched it slowly creep up to 2.5V. (Took a several hours.) Then I disconnected it completely. I've been periodically checking it now for some time.

@ 0 hours, 2.490V. (Just disconnected from the charging circuit.)
@ 8 hours, 1.936V.
@ 19 hours, 1.685V.
@ 34 hours, 1.640V.
@ 44 hours, 1.481V.

Q: Do supercaps really drop that much voltage over a ~48 hr period?

[Flávio V]

The supercap should have retained near 2.2V in those 44H....when i charge my 350F maxwell ones after 8 hours they go from 2.7V to near 2.5-2.6V...so those super caps might be faulty or it is normal leakage...

Ioxus can leak at 1mA...the maxwell ones leak at 300uA(max or both)

[Nickk2057]

hehe.... did it go boom yet?

anyways.. if it hasn't... congratz

[chris.jarboe]

Samurai,

The initial leakage current for a supercap is typically high. The datasheet for the Ioxus part specifies 1.2mA for the leakage current, but this is "72hr leakage," or the leakage current after being charged for 72 hours. So yes, this is normal when they are new.

As for your original questions:
1) Absolutely correct.
2) Absolutely correct.
3) Ideally you want as little series resistance as possible, so I would start without it.

Good Luck!

[SoftwareSamurai]

The initial leakage current for a supercap is typically high. The datasheet for the Ioxus part specifies 1.2mA for the leakage current, but this is "72hr leakage," or the leakage current after being charged for 72 hours. So yes, this is normal when they are new.

Thanks Chris.
So, when you say this is normal when they are new, do you mean that it will leak less after many charge/discharge cycles? (i.e. As it gets "older"?)

[chris.jarboe]

Yes. The leakage current will eventually be what you expect from the datasheet.

[SoftwareSamurai]

Interesting. That's not what I would expect from a capacitor.

I did contact IOXUS and they would like me to repeat my experiment and take measurements every 8 hours to double-check the discharge rate. Looks like I know what I'll be doing next weekend!

[jh15]

I bought a Coleman capacitor screwdriver just to have the concept in hand of capacitor power.

If I did not buy this product, I would think it was all vaporware. Great to demo to people. I drive many  screws, then complain, oh gee my screwdriver needs to be recharged... they expect it to take hours, it takes about 30 seconds.

[Psi]

Then I disconnected it completely. I've been periodically checking it now for some time.

@ 0 hours, 2.490V. (Just disconnected from the charging circuit.)
@ 8 hours, 1.936V.
@ 19 hours, 1.685V.
@ 34 hours, 1.640V.
@ 44 hours, 1.481V.

Q: Do supercaps really drop that much voltage over a ~48 hr period?

No, that's not normal.
My 2.5V maxwell 2600F caps take 1-2 months to drop from 2.5V to 2.0V and after that the loss is much slower.
I've got one here which hasn't been charged in ~6 months and its still at 1.8V

You either had something connected to the caps which was drawing current or your caps are faulty (perhaps small internal leakage).