Maker Pro
Maker Pro

Finding a nano amp transistor for Solar LED Light Marker

I'm looking to build a circuit for a Solar LED light Marker using a very low powered LED like a 0.06W 3528 SMD LED or even lower wattage as the light is to denote where a path is, not to light up a path.

I found a useful circuit at http://www.reuk.co.uk/wordpress/solar/use-solar-panel-as-darkness-detector/, but I wanted to use a 3V 50mAh lithium rechargeable battery and a 52mm x 19mm solette rated at 0.5V and 0.14 watt.
I have measured output from solettte on a multi meter and in dark conditions it produces 0.003mA at 9mV which is about the level or lower that I want the LED to come on.

How would I amend the circuit for this, in particular the transistor, as I cannot find a transistor that will switch at such low amps (3nA !) and voltage?

Thanks

Mike
 
I'm interested in electronics and I thought it would be a good project to work on and I already have several Solettes I can use. So just trying to understand how to build a circuit that will not switch the LED on until the current and voltage are VERY low as I don't want to switch LED on until is really quite dark.
 
I have many cheap Chinese solar garden lights.
The old ones used solar panels and light dependant resistors that got sunburned, and old Ni-Cad battery cells that failed or rusted away in a couple of months. Newer ones use a solar panel with a glass cover that does not sunburn and it detects the light. The battery cell is a modern Ni-MH that is in a case that does not rust away. They cost almost the same as before. But in winter there is not enough strong sunlight to fully charge the battery so the LED lights for only 1 hour. If the daytime is cloudy then in winter the LED does not light at night.
 
Just my 2c.
Why have them turn off during the day if they are to just mark a pathway?
It adds complication to the circuit, the solar panel has more than enough output to still recharge the battery with the led lit and no one will notice that the led is on during daylight.
 
What is going to charge your 3V battery? Not the 0.5V solette.
You made a typo, in dark conditions does the solette produce 0.003mA (3μA) or 3nA?
The 9mV is not nearly enough to switch on a transistor that needs about 67 times more. I think you want more than 9mV to turn on the first transistor that turns off the second transistor that drives the LED.

You show you are in the UK that is farther North than me here in the south of Canada. Now that it is the middle of winter the daylight hours are low and the sun is low in the sky and many days are cloudy. Then my solar garden lights barely light for 1 hour at night (they light all night in summer), and they use a much larger battery and solar panel than you are planning. But your LED will be extremely dim so if you find a way to charge your little 3V battery then its charge might last for one hour at night in winter.
 
I think a joule-thief charger circuit may be helpful here.
Use it to charge your battery at day light time(boost the solette voltage).
Than the circuit in #1(with the same solette to control the LED activation at dark) may be used.

It is worth testing the following approach for the LED lighting:

Instead of constant current through the LED , use a very short "spike" of higher current through the LED.
e.g. 10msec (LED on) once every 2 seconds.(that is a ratio of 1:200)
The setting of this needs experimenting to get the desired optimal working condition for
"clarity of the path" and lower power consumption.
 
Thanks for replies:
Yes there is mistake - in dark conditions, the solette produces 3uA at 9mV, not 3nA, and the battery should be 500mAh at 3.7V, not 50mAh at 3V.
I am a newbie at electronics, so this is more of a project to learn electronics which is semi-useful, so not looking to buy a product if one existed, but on the practical side, I live in a village in the UK that has no street lighting, so when there is no moon you cannot literally see your hand in front of your face, so even a very dim led would be enough to show where a path is.
So as I understand a 3.7V needs to be charged at 4.2V so the power required to charge 500mAh would be
0.5*4.2 = 2.1w, so if the solette was charging at 0.14w then it would take 15 hours to charge, so battery is more than big enough, so supposing I got half the output of the solette over 7 hours, then I would get about 0.5w of power stored in the battery so this in theory can power my 0.06 LED for 8 hours.
But as Audioguru pointed out, in the winter you get far less sun, so right now at about midday at temp of 5C, when the sun is behind the clouds I get 20mA at 0.55V and when the sun is out I get 100mA, so I might only be able to power my LED for a few hours, so I don't want it to come on when it is a bit dark (like my bought garden LED lights do) as this is pointless as the LED is "to be seen", not to light up an area and he battery will have run out of charge before it gets properly dark, when I actually need the LED.
So I am aware there are a few issues:
  1. I need to step up the 0.5V from solette to 4.2V to charge battery
  2. I wan't the LED to come on as dark as possible and therefore I need a transistor or other electronic component that works with very weak signals
So do such transistors exist and if not then what electrical component would I use?

Thanks

Mike
 
A transistor or "joule thief circuit" will not work until its supply voltage is at least 0.65V. The output current from a joule thief circuit is very low. If you use two solettes (what is a solette?) in series to produce 1.1V then when the input of a voltage boosting circuit is 20mA its output current is (1.1V x 20mA)/4.2V= 5.2mA x 0.8 for efficiency loss= 4mA. At 4mA your 500mAh battery will take 500/4= 125 hours/24= 5.2 full days or 125/8 winter daylight= more than two weeks to fully charge.

Here is the datasheet for the IC in a solar garden light. It usually charges and is driven by a 1.2V Ni-MH rechargeable battery cell and continues to produce light in the LED until the battery has dropped to 0.9V. Its light sensitivity is adjustable and its LED current is determined by the value of its inductor. If you use a huge value inductor then the LED current will be as low as you want. The solar panel is about 2V (4 cells) at 30mA to 40mA but can be smaller since your LED current will be very low.

Maybe you can use a second IC to boost the voltage of your solette to 2V.
 

Attachments

  • pr4403.pdf
    198.8 KB · Views: 70
You would need a comparator with rail-to-rail inputs and outputs. If your LED is on the order of 20mA or less, you could drive the LED directly from the comparator. The booster would have to be turned on by another transistor from the comparator output.

Bob
 
Audioguru ,
A "solette" is a nickname for a small part of a solar cell/panel.

A "joule thief" can be made with a Germanium transistor(like this one) it will work well below 0.5V input voltage.


The idea was not to have a constantly lit LED at night but to pulse it on/off at a ratio of 1:200 (or so,as described above.),that would cut the consumption a great deal.
 
The solettes I have are bare polycrystalline solar cells which have no wires attached, so you can actually break them into 2 or several pieces, but they are very thin (0.2mm) and hard, so quite brittle and very hard to break into nice regular pieces, but in theory I could split a singe solette into 8 pieces, so then I could get over 4V as I actually get slightly over 0.5V from them in the daytime, even with cloud.
So for charging I guess I have few methods to boost the 0.5V:
  1. Splitting solette into smaller pieces and placing them in series
  2. Using a "joule thief"
  3. Using a boost voltage regulator
  4. Maybe using comparators - not quite sure how this would work.
But still have problem switching at low voltage and current - seems a Germanium transistor will help here, but don't really understand the spec sheet (https://www.alliedelec.com/m/d/fbd71dce0288341be24e97741da8001a.pdf) as I am new to electronics.

As I understand, the standard switching voltage for a transistor is 0.7V, but I think the Germanium transistor maybe switches at 0.2V, but I am not sure if the value of the base voltage and current at which the transistor switches (allows current to flow from collector to emitter) is dependent on the collector voltage and current, so then if I was to follow the circuit wiring at http://www.reuk.co.uk/wordpress/solar/use-solar-panel-as-darkness-detector/ with my 20mA 3.6V 3528 SMD LED connected to the emitter, then my collector voltage and current would need to be the same 20mA at 3.6V, but if a lower voltage and current for the collector means I can switch with a weaker signal at the base, then I guess I could connect the emitter of the transistor to the base of a second transistor and connect my LED to the 2nd transistor, so that I could use a much lower voltage and current for the collector of the first transistor which is connected to the weak signals from my solettes.

Basically I want the LED to switch on as late as possible when it starts getting darker, meaning I want to switch at the lowest possible voltage and current from the solette, but not at all sure what the minimum voltage and current I can achieve is.
 
The LED and its current-limiting resistor connects to the collector (not emitter) of the PNP transistor. The emitter connects to a voltage higher than the LED so about 5V for your 3.6V LED. I did not know that old germanium transistors are still available so its input must be about 0.2V and another 0.2V for its base current limiting resistor. ANY transistor will do nothing if you feed it only 0.009V. To make the required 0.4V you will need 0.4/0.009= 44 or 45 solette cells in series but the transistor will be destroyed by the high power from them in sunlight.

An old fashioned light dependant resistor can be used to turn on a transistor but they get sunburned and don't last long in sunlight. A photodiode needs a fairly bright light for it to do anything and a photo-transistor is better but not for your very low level of light.
 
The comparator does not help you with the low voltage for charging, it allows you to switch the LED with a low voltage. Yes, even 9mV.

Bob
 
Top