Go Down

Topic: 4 Legged capacitor?! (Read 12302 times) previous topic - next topic

dc42

The issue is that you are driving L2 from a low-impedance source, so you have very little control over the current that flows through it. You are likely to either fail to light the LED up at all (because its forward voltage is higher than 2.7V or whatever voltage you charge the capacitor to), or burn it out due to excessive current.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

bigred1212

#31
Aug 13, 2013, 08:07 pm Last Edit: Aug 13, 2013, 08:19 pm by bigred1212 Reason: 1
I will use just a standard red LED with no specs for L2, so I am assuming 2v forward drop and 20 mA.  With a 2.7v source in the capacitor, the online calculator suggest a 39 ohm resistor in series.  

I kind of looked at that and said 1)2.7 v vs 2v, meh;  2) 39 ohms is practically a wire; and, 3) I don't have one of those anyway.  

I'll go to the Shack and see if they have a horribly overpriced package of 5 smallish resistors and stick one in there to do SOME current limitation.

Would you concur with a smallish 39 ohm value or suggest something else?  Or would you suggest a completely different testing methodology?  All I really want to do is see if I could use a giant cap (instead of a rechargeable battery) to drive a solar garden light.

EDIT:  In general, what is the preferred method of driving a load needing current limitation from a low impedance source? Particularly when it is a low impedance, low voltage source for which voltage dropout due to something like a diode or resistor means there could be insufficient voltage to accomplish the desired work?  Some sort of shunt regulator?


dc42

#32
Aug 13, 2013, 08:18 pm Last Edit: Aug 13, 2013, 08:20 pm by dc42 Reason: 1

39 ohms is practically a wire


Not at all, a wire is more like 0.01 ohms. I agree that if you will be charging the capacitor to its maximum rating of 2.7V, then 39 ohms or 33 ohms is about right.

To avoid buying overpriced packets of resistors, consider investing in something like this http://www.ebay.co.uk/itm/640-Pcs-64-Values-1R-10MR-1-4w-1-Metal-Film-Resistor-Kit-Assortment-Pack-Mix-/300925197516?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item46108a1ccc.

btw solar lights normally use a boost converter IC so that they can work down to very low voltages.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

dc42


EDIT:  In general, what is the preferred method of driving a load needing current limitation from a low impedance source? Particularly when it is a low impedance, low voltage source for which voltage dropout due to something like a diode or resistor means there could be insufficient voltage to accomplish the desired work?  Some sort of shunt regulator?


Either a constant current linear series regulator, or a constant-current switching regulator. A constant current switching regulator works well for small LEDs, because they don't need a continuous current, they can work on pulses of current provided that the amplitude of the current pulse is controlled.

You might want to look up "joule thief". This is an example of a switching regulator with low efficiency compared to modern ICs, but it works down to very low voltages.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

bigred1212

Yeah, I was wondering about a joule thief. 

So, if I made the capacitor in my original circuit the power source for this:



which is straight from
http://en.wikipedia.org/wiki/Joule_thief

do you think that might have more success? 

Or would you have a particular higher efficiency constant-current switching regulator IC that you would recommend?
Maybe something like the 756 version of this:
http://www.digikey.com/product-detail/en/MAX756CPA%2B/MAX756CPA%2B-ND/1130177
??

dc42


Yeah, I was wondering about a joule thief. 

So, if I made the capacitor in my original circuit the power source for this:
do you think that might have more success? 


That will work down to a lower voltage. However, it's quite hard to control the LED current using that circuit, it depends on all 4 components. If the LED you are using has a forward voltage less than about 3V then you need to connect the LED cathode to the positive supply instead of to ground.


Or would you have a particular higher efficiency constant-current switching regulator IC that you would recommend?
Maybe something like the 756 version of this:
http://www.digikey.com/product-detail/en/MAX756CPA%2B/MAX756CPA%2B-ND/1130177
??


This http://uk.mouser.com/Search/ProductDetail.aspx?qs=nSryOFbzj8L3ESkNRVFoVQ%3D%3D is far more suitable.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

bigred1212

I don't understand the first part of that.


As for the chip recommendation, thanks for that.  That little 5 legged dude looks like the chips I see in commercial garden lights.

if I can figure this out (and there is absolutely NO guarantee I can), I might have a circuit with a 4 legged cap and a 5 legged chip. I find that amusing.

Go Up