Charging a capacitor with solar panel?

[Artlav]

A solar panel have a certain voltage at which it produces the most power.
Below this voltage it's a current source - giving about the same current all the way down to short circuit.
Right?

On the other hand, a capacitor sucks up as much current as is available (ignoring ESR), slowly climbing in voltage.

So, if i attach a big capacitor (12F supercap) to a solar panel, it would be charging it with the efficiency of a linear regulator from it's optimal voltage down to about zero.
Right?

If so, can it be improved?
What kind of circuitry can accommodate a capacitor charging at high efficiency from a solar panel?
Some sort of current limiting buck converter that tracks it's input voltage rather than output one?

Are there any circuits like that or a googleable name for one?
Specs in my case are 17V, 0.5A peak panel, 12F 22V supercap.

 

[phinds]

On the other hand, a capacitor sucks up as much current as is available (ignoring ESR), slowly climbing in voltage.
.

No, it does not. I DOES climb in voltage but "sucks up" only the amount of current required to get it up to the applied voltage level. How much current that is depends on the size of the cap.

That is, if you had a source that could supply 10volts at 1 million amps and you applied it to a small cap, hardly any current would be drawn. Compared to the million amps available, the amount drawn would be approximately zero.

 

[Averagesupernova]

That is kind of nitpicking isn't it phinds? Saying that the voltage climbs implies that it will not draw as much as is available forever.

 

[phinds]

That is kind of nitpicking isn't it phinds? Saying that the voltage climbs implies that it will not draw as much as is available forever.

Doesn't seem like nitpicking based on the OP's misunderstanding.

 

[Artlav]

I was talking in practical terms - while charging, a large capacitor continuously trying to draw much more current than the solar panel can provide, almost to the end of the charging.
So, the voltage on the panel goes down almost to the current voltage on the capacitor, which for the most time is way below it's optimal voltage. Thus, it delivers much less power than it could.

Is that a correct description?
Generally asking, is there a way to charge a capacitor from a solar panel with more efficiency than just connection one to the other?

 

[phinds]

If the voltage on the solar panel goes down the when it hits the voltage to which the capacitor has charged, there will be no further charging, and unless you have a diode to prevent it, the cap will start to discharge into the panel.

Your whole scenario is confusing. It would be more clear if you would use specific numbers, but it still sounds to me like you don't understand what's going to happen.

 

[Averagesupernova]

I don't think you are getting it phinds. The only thing in the OPs scenario that is bringing the voltage of the solar panel down is the fact that it is loaded by the capacitor. As the capacitor charges, and it has to if it is passing a current, the voltage comes up. Eventually the capacitor will charge to the voltage of the unloaded solar panel.
-
Based on the following quote what appears to me is that the OP wants to know whether there is a more efficient method to connect the solar panel and the capacitor.

So, if i attach a big capacitor (12F supercap) to a solar panel, it would be charging it with the efficiency of a linear regulator from it's optimal voltage down to about zero.
Right?

A lot of voltage is being dropped across the internal resistance of the solar panel when the capacitor is first connected. Lost power here. The OP is correct in thinking of the loss as being similar to a series pass transistor type regulator. So what's the fix? Charge the cap with a switching supply of some kind. Make it go into current limit if the current demand exceeds what the solar panel can source. It will at that point act as a current source instead of a voltage source. If the losses in the switcher are not enough to exceed the losses when doing it the old way, the cap should charge faster.

 

[OmCheeto]

...
Generally asking, is there a way to charge a capacitor from a solar panel with more efficiency than just connection one to the other?

My guess is no.

And why do solar panel questions always pop up when it's cloudy outside?

And why don't I have any super/ultra-capictors in my toolbox yet?

Anyways, on a somewhat related side note. I saw a solar powered tack welder this year at the local fair.

There was no intermediate circuitry.

Two super-caps, one solar panel, one knife switch, and some sheet metal scraps.

It was beautiful.

 

[Artlav]

The thing in question is a charmingly straightforward solar powered phone charger.

The phone does not like brownouts when clouds cover cover the sun - it throttles the charge current down, but never up.
So, i added a 2kJ (3% of the phone's battery worth, or about 5 minutes) supercapacitor bank to ride through the clouds.
The problem is, if the bank gets empty enough (below 10V or so) it takes a whole lot of a while to recover.

The first 10V on the capacitor is 25% of the energy, the second 10V is the other 75%, but the first 10V takes a whole lot of a while longer to achieve than the second 10V when it charges up from flat in the morning.

So, i was thinking - is there a way to improve the efficiency on the panel->capacitor side?
What would that switching supply looks like and what would it be called?

Hope that clears up the problem.

 

[OmCheeto]

I don't think you are getting it phinds. The only thing in the OPs scenario that is bringing the voltage of the solar panel down is the fact that it is loaded by the capacitor. As the capacitor charges, and it has to if it is passing a current, the voltage comes up. Eventually the capacitor will charge to the voltage of the unloaded solar panel.
-
Based on the following quote what appears to me is that the OP wants to know whether there is a more efficient method to connect the solar panel and the capacitor.



A lot of voltage is being dropped across the internal resistance of the solar panel when the capacitor is first connected. Lost power here. The OP is correct in thinking of the loss as being similar to a series pass transistor type regulator. So what's the fix? Charge the cap with a switching supply of some kind. Make it go into current limit if the current demand exceeds what the solar panel can source. It will at that point act as a current source instead of a voltage source. If the losses in the switcher are not enough to exceed the losses when doing it the old way, the cap should charge faster.

Ah ha! I think I see where you and the OP are going here.

I'm glad I said I was just guessing.

 

[The Electrician]

Are there any circuits like that or a googleable name for one?

Look up the phrase "maximum power point tracking", (MPPT):

http://en.wikipedia.org/wiki/Maximum_power_point_tracker

 

[phinds]

I don't think you are getting it phinds. The only thing in the OPs scenario that is bringing the voltage of the solar panel down is the fact that it is loaded by the capacitor. As the capacitor charges, and it has to if it is passing a current, the voltage comes up. Eventually the capacitor will charge to the voltage of the unloaded solar panel.
-

The way I read it was that the OP wanted to consider what happens when the voltage on the solar panel is on a downward slope. My comments were entirely based on that so if I misunderstood that, then yep I got it wrong.

 

[OmCheeto]

The way I read it was that the OP wanted to consider what happens when the voltage on the solar panel is on a downward slope. My comments were entirely based on that so if I misunderstood that, then yep I got it wrong.

It's Saturday. Nothing wrong with guessing wrong on a Saturday.

And solar panels are freakin' non-linear. Gosh-darned semi-conductors...

They've always made my brain hurt.

So who is going to do the math, and figure this out?

 

[meBigGuy]

I am assuming you want a simple method that provides optimum power transfer when the panels are fully illuminated, and are not so worried about tracking optimum power point over varying conditions. So, basically, you want an interface between the cells and the supercap that presents a contant voltage to the cells and a charging current to the supercap, up to the cell voltage.

Such a thing would be called a solar panel supercap charger, and I just found one (after I wrote the top paragraph to get focused).

http://circuits.linear.com/339

 

[ScienceGeyser]

I suggest putting a low voltage disconnect circuit on it and just never letting it discharge completely. Since you will loose a significant amount of power when operating at lower voltages with the solar panel, you should simply make the circuit turn off until there is adequate voltage to work in an efficient range. A low power microcontroller could be used or it could be possible to use an LM555 to make a circuit that turns off when the voltage drops below a certain voltage and stays off until you reach some higher voltage. The LM555's are great little circuits and don't have to be used as timers. Briefly, you could connect pins 2&6 to the +cap terminal (with some resistor network max 16V on the 555) and use pin 5 to control the turn on voltage and then use the output to control your switch to the buck converter. If the logic level is wrong for the control you choose, you can use the open collector discharge pin and a pull up resistor to get the opposite logic level. This may be a good starting point if you don't mind that you can't get above 1/2 the applied voltage for the turn on voltage. You will need some other support circuitry and this solution is not super simple, but it can keep you in the most efficient operating range of your solar panel.

 

[meBigGuy]

?? Why do all that when there is a chip designed to do exactly what he wants to do?

 

[Artlav]

Look up the phrase "maximum power point tracking", (MPPT):

I suggest putting a low voltage disconnect circuit on it and just never letting it discharge completely.

Sounds interesting, thank you.
Just disconnecting at something like half the bank's energy capacity should be quite effective at removing the problem, compared to implementing MPPT.

?? Why do all that when there is a chip designed to do exactly what he wants to do?

Because A. anything by LT is high unobtanium here, and B. That chip can only do 14V:
"The charger employs a 3.3V float voltage feedback reference, so any desired battery float voltage up to 14.4V can be programmed with a resistor divider.", which is a bit too low.
Although, the existence of such ICs do provide a new direction for searching.

 

[meBigGuy]

OK --- I need to watch my attitude :).

A general approach is to build a voltage regulator that regulates the panel to a constant voltage by supplying current to the supercap. You probably mostly only care about being efficient when the panel is being lit well. So, you can choose a fixed voltage to regulate to. You probably want to use a buck switching regulator so you don't waste power due to the voltage drop from the panel to the supercap.

Run the solar panel into a moderate sized capacitor and then a regulator to the supercap. Just to prove feasibility try try this with an LM317. You need to connect the LM317 reference input to the panel so the LM317 will try to keep the input voltage constant. But ...
The trick is to invert the feedback so the regulator will supply less current if the INPUT voltage drops. The LM317 want to see 1.2V. Normally it would pass more current if the reference was low. But, you inverted it with an opamp so it draws more current when the panel voltage is high.

The LM317 will dissipate power, (I didn't do the math) so you want to replace it with a switching converter.