Jule thief and super capacitor circuit idea

[Fluffyboii]

Hi,
Recently I came across a electronic store that sells supercaps in various sizes. There were going up to 2500F I think with one being the size of a Pepsi can but highest capacitance to price ratio ones were I think the 200F ones. Anyway I bought a 5F one to play with since it was about 1USD. I also bought 2 1.5V solar panels to see how long it would take to charge this 5F cap with.
First of all I connected it to the joule thief I build some time ago out of boredom. I put a 1W led with voltage drop about 2.8-3V to it. When fully charged to 2.7V it runs the LED about 3 minutes and starts getting weaker after 0.6V and cuts off after 0.5V. Which is impressive to be honest considering its small size.
I am not sure about the schematic of the joule thief but it is the classic one using 1K ohm resistor at base of the 2n3904.
With two cheap solar cells connected in series and a germanium diode with voltage drop of 0.25V It took about a long time to charge it under my desk lamp. It got stuck around 2.5V at some point so I used another florescent lamp instead of a led one and placed it closer. One issue was that capacitor got charged about 2.9V at some point so I immediately discharged it and it seems it is not damaged. Perhaps using a silicon diode might be better to have more voltage drop or two series schottky diodes may barely keep it under its maximum voltage limit.

I have few questions: How can I make a simple thing to limit the charging process at 2.7V to keep the cap safe. Looks like these 1.5V solar panels can go up to 3.3V when connected in series under a strong light. I couln't test them under the direct sun yet so I am not sure how they would work under it. Using a diode with higher voltage drop will decrease efficiency so is there any other way to make a simple device that cuts off after a certain threshold. Lets say I want to make one of those shake to charge flashlight things by using a solenoid and passing a magnet through it instead of solar cells because lets be honest charging a flashlight with solar cells doesn't make much sense. The voltage produced by such a setup would be quite unstable and can peak quite high. Using complicated circuit would eat away efficiency. Perhaps there is simple solution for this?
Maybe two supercaps in series to get 5.4 max voltage and using a 5V solar cell would make more sense but not sure how joule thief will work if source voltage is higher than LED Vd.

Can these caps handle being charged more than what they are rated for. I mean normally it is nice to have them under the max limit but considering how little voltage these can handle it makes sense to charge them as much as possible.

Finally I would like to make the joule thief more efficient if possible. There is about %20 of power left in the capacitor when LED turns off. If I use a smaller torroid to increase its operating frequency, would that make it more efficient.

I am not sure what I am going to do with this. My curiosity got better of me.

 

[Nigel Goodwin]

I couln't test them under the direct sun yet so I am not sure how they would work under it.

Direct sunlight is MASSIVELY more effective than artificial light - if you've ever watched the 'Robot Olympics' there's a solar model 'car' race (sprint) - done indoors under an absolutely huge array of specialised lighting. It looks like if you walked underneath them you'd be fried to a crisp

Can these caps handle being charged more than what they are rated for. I mean normally it is nice to have them under the max limit but considering how little voltage these can handle it makes sense to charge them as much as possible.

Completely the opposite, they are VERY sensitive to over voltage, and very prone to failure as well. The few VCR manufcaturers who were stupid enough to use them had massive failure rates - still, good for me, who got paid for repairing them

I really hate 'super' caps - they are so badly named - it's a poor imitation of a battery, so why not just use a battery?.

 

[Fluffyboii]

Direct sunlight is MASSIVELY more effective than artificial light - if you've ever watched the 'Robot Olympics' there's a solar model 'car' race (sprint) - done indoors under an absolutely huge array of specialised lighting. It looks like if you walked underneath them you'd be fried to a crisp



Completely the opposite, they are VERY sensitive to over voltage, and very prone to failure as well. The few VCR manufcaturers who were stupid enough to use them had massive failure rates - still, good for me, who got paid for repairing them

I really hate 'super' caps - they are so badly named - it's a poor imitation of a battery, so why not just use a battery?.

I assumed these points. Looks like putting it under sunlight as it is would obliterate the poor capacitor. Considering %10 over the threshold is not good for caps in general these having low voltage makes it too easy to overcharge them. I also heard about old electronics self destructing overtime because some companies tried using these caps as a replacement for coin cell batteries to hold basic info in memory like real time clock etc as you said.

Using a regular Ni-Cd battery would be no fun though
I had a shake flashlight I lost long ago that used a large cap to work in a similar manner, just got interested what could be a good use case for these things since they don't really makes sense because of the poor size to energy capacity ratio and price. But at least no worries of amount of charge cycles or safety risks right?

 

[rjenkinsgb]

Try four 1N4148 or similar in series, across the cap? That should limit it to around 2.4 - 2.6V hopefully

Ideally keep the voltage on the cap to no more than 2.5V, to get a decent life from it.