A friend has come into 7 solar panels, each meant to power a string of 50 LEDs. We have none of the lights, and I'm wondering what output to expect from the device. I'm guessing it's a high frequency, boosted voltage, low duty cycle but I'm hoping somebody here has seen one of these before.

The housing of each panel contains a 4v, 1AH battery, circuitry on a small PCB, and a CdS cell, along with an on/off switch. The panels are about 4x5" and produce over 5v and at least 0.1 watts peak power in full sunlight. These solar light sets were made by International Development (IDC), a subsidiary of Westinghouse.

 

I'm guessing it's a high frequency, boosted voltage, low duty cycle but I'm hoping somebody here has seen one of these before.

Solar cells are DC. So it won't be a high frequency.

 

Thats correct. Unless there is a built in controller, you will get straight DC out of the panel.

These devices sometimes use a typical SEPIC circuit to boost the voltage. There are many IC's now that handle the low solar output rather nicely to buck/boost the voltage to a useable level.

I developed a little design that used a single .5v 3.3A solar cell to run a uC.

The uC simply recorded how long is was on. Therefore I had a cheap data logger that would let me know how much sun fell on a particular area, if shade affected the area and what not.

A comparator watched the voltage. If it dropped below 4.2, it shut down. This allowed the uC to write to the flash properly. If it dropped to low, I would get improper data.

SO, what does all this mean? You can get a usable voltage from the smaller panels or cell with a little power management and storage. Storage can be a supercap or battery.

If space is an issue, the super cap is a good way to keep a nanoWatt type uC happy during the low-sun times.

A comparator can trigger an interrupt to put the uC in sleep mode until a circuit, or even a zener can wake her up when proper levels are reached.

 

Solar cells are DC. So it won't be a high frequency.

Well, the ICs you can buy to boost voltage do so at 80kHz, putting out a ~10% duty cycle spike at that frequency. The LEDs appear to be constantly on at that frequency since it's about 2000X faster than your eye can detect. But I don't know if these Westinghouse things use that approach.

I guess everything is relative, but to me, any frequency over 60Hz is "high".

 

Thats correct. Unless there is a built in controller, you will get straight DC out of the panel.

These devices sometimes use a typical SEPIC circuit to boost the voltage. There are many IC's now that handle the low solar output rather nicely to buck/boost the voltage to a useable level.

Right, I should have been more clear: I have no question about what the panel itself can do, but rather what to expect coming out of the entire unit to the LED string (which I don't have). And how to evaluate that output. Do I need an oscilloscope? Maybe I'll hook up a diode and a cap to at least give me a peak detector.

BTW, thanks for the project idea. That's where this is leading. Once I know what these things are doing, I'll help the young man that owns them to pick a project for them, and then see it through. I think it'd be a fun and educational experience for him to actually build something.

Know of any other good ideas for these panels? Since they're so big, I'm more interested in power-related projects instead of light-sensor projects. They'd be overkill for merely detecting light. I'd rather drive some lights or a motor or something, charge a cellphone maybe.

 

Do you mean 100 mA output rather than .1W? A 4X5 cell should put out around 1/2 W. Try the peak detector for output V. Also try loading the output with a resistor so as to draw around 20 mA, if V falls only a bit try 40 mA; strings may be series parallel.

 

Do you mean 100 mA output rather than .1W? A 4X5 cell should put out around 1/2 W.

The 0.1 watt peak was measured by covering the CdS cell and loading the output of the entire device - not just the panel - with various resistors. Peak power was with a load of ~100Ω, at about 4v. This was measured BEFORE I opened the thing up and discovered the (dead) battery, which of course would have been adding to the load. So the panel itself was likely making more than the 0.1W I saw at the output. Actual dimensions of panel are 3.5"x3.75", or 13.125 sq. in.

Hmmm. I saw a DC voltage across the resistive load. Duh, I guess it's not putting out AC. Maybe pulsed DC. I'll check the peak voltage.

[update] Huh, the peak voltage is just the same as measured directly, less the diode drop of the peak detector. So it does NOT seem to be pulsing a high voltage. I guess this means the LEDs were wired in parallel. If the circuit pulses at all, it's to boost current and/or limit duty cycle, not to boost voltage. I can post a photo of the PCB if that would help. It must be about duty cycle: You couldn't run 50 LEDs at 5-10mA for 10 hrs with one 4v, 1AH battery.

[update 2] Looks like I'm talking to myself here! Anyway, I just hooked up a 12v light bulb to the output and the voltage dropped to 0.22v, so no light. The battery by itself will light this bulb to an orange glow, so clearly the circuit is not just channeling battery power out to the LED string. Then I tried two LEDs in serial. No light. Two LEDs in parallel look great, so apparently this thing is meant to drive a string of LEDs in parallel.

I guess that's about all I need to know for now, although I am still a little curious what the PCB is doing to the output. I'm guessing PWM but I don't have an easy way to check that.

 

Talking to yourself is good. Are you answering yourself as well???? Are they the right answers???

Posting a photo would be helpfull.

All of the xmas lights I have that do fancy things are generally controlled by a small circuit consisting of an IC (OEM blob) on a PCB with a few diodes and resistors. They usually have a switch to cycle through the different displays. Most have 4 outputs with every forth LED connected to each output. What this chip does I have no idea other than keep the Minister for War and Home Affairs happy.

To see exactly what your box is doing you would need a CRO.

 

Talking to yourself is good. Are you answering yourself as well???? Are they the right answers???

I am! ...and I have no idea.

Posting a photo would be helpful.

I took some but I'm pretty satisfied with my testing. I'm not investing more effort until the owner expresses interest in a project. In words, the front of the PCB has one transistor, a diode or two, at least one cap and resistor. The back has a SMD IC with maybe 14 or 16 pins.

All of the xmas lights I have that do fancy things are generally controlled by a small circuit consisting of an IC (OEM blob) on a PCB with a few diodes and resistors. They usually have a switch to cycle through the different displays. Most have 4 outputs with every forth LED connected to each output.

This one isn't so fancy. Just two wires carry the output, which I believe is just full on all the time.

 

My solar garden lights have a solar panel that is half the voltage of yours and is half the size. They charge a single 600mAh Ni-Cad cell that powers a voltage stepup circuit that drives a single LED for a few hours at night. The LED is visible in the dark but it is not bright and it doesn't light-up anything.
So your 50 LEDs must be extremely dim since they share the low current if they are on for a few hours.

 

My solar garden lights have a solar panel that is half the voltage of yours and is half the size. They charge a single 600mAh Ni-Cad cell that powers a voltage stepup circuit that drives a single LED for a few hours at night. The LED is visible in the dark but it is not bright and it doesn't light-up anything.
So your 50 LEDs must be extremely dim since they share the low current if they are on for a few hours.

I never saw them in action, but I had the same question. The battery is only 1AH, albeit 4v, so a few more watt-hours. But still... The sales claim is 6 hrs at night. Hmmm..., maybe when they're brand new and get a full charge.

482101MT, available new from several sources:
http://www.google.com/search?q=4821...esult_group&ct=title&resnum=1&ved=0CCsQrQQwAA

http://doitbest.com/Christmas+light...opment-model-482101WH-doitbest-sku-904908.dib

 

I could not find out any details for the solar light except it is turned off after 6 hours with a timer.
Its battery type is not listed but if it is capable of 1/6th of 1A fo 6 hours then the current is 167mA and each of the 50 LEDs uses a current of only 3.34mA if they are in parallel. Then they are quite dim.

 

I could not find out any details for the solar light except it is turned off after 6 hours with a timer.
Its battery type is not listed but if it is capable of 1/6th of 1A fo 6 hours then the current is 167mA and each of the 50 LEDs uses a current of only 3.34mA if they are in parallel. Then they are quite dim.

What is the effect on perceived brightness of pulsing a higher-than-average current at a lower-than-average duty cycle? Can you make them seem brighter using that strategy? The output from the panel is definitely NOT just a pass-thru of the battery, it won't even glow an auto light that the battery is capable of lighting.

 

Your vision reacts to brightness fairly slowly so it averages a quickly-pulsing light then it appears dimmed. Also, an LED has an increased loss at higher currents.

The 4V battery can easily light some low current dim 3.5V LEDs but not a 12V high power light bulb.