After two months of use, I noticed the LEDs weren't as bright, even though there's been plenty of sunlight.

I opened up the casing and measured the 18650 battery inside at 4.9 volt. This seemed unsafe, so I sent it to the recycling bin and meanwhile I've been trying to understand the circuit, so I can keep using it in a safely manner.

Please note the photos in attachment:

- solar panel outputs close to 6V under direct sunlight;
- if the solar panel isnt plugged in, the LEDs light up whenever there is movement; when it's plugged it, the LEDs only light up at night.

My conclusion is that PIN6 is pulled High when the solar panel is connected, disabling the PIR but charging the battery through the transistor.

If the panel is at 6V, is the 18650 exposed to those 6V and will take as much current as the solar panel can provide, considering the base of the transistor is biased by the 100 ohm and 10kohm resistors?

Also, the LEDs are exposed to over 4.2V, correct? They pulled 600mA at 3.3V under test...

Regarding making charging safer and LEDs last longer, can I connect the emitter to the input of a TP4056 board to charge the battery (output to B+ L+)? Can I solder resistor to the anode of the LEDs to limit the current, when B+ is at 4.2V?

 

The uC (microcontroller) is monitoring the battery voltage at pin 1, and controlling charging at pin 6, possibly with PWM (pulse-width modulation). It also might be controlling the lights with PWM at pin 5.

The PWM possibilities mean that measuring voltages with a meter might not be accurate. Only a scope will show the true waveform. If you don't have one, there are very low cost scope modules and kits on ebay.

OTOH these are ridiculously low cost devices, and they might in fact be overcharging the pants off of the battery and overdriving the LEDs. What are you going to do about, complain to the manager?

ak

 

Put together one of those soundcard scopes, there's a square wave coming out of pin5, powering the LED through the mosfet.

 

Under full sunlight, voltage created by the solar panel is 6.5V and short circuit current is 135mA.

But how can I prevent the 18650 to charge over 4.2V, considering the one I removed was at 4.9V?

 

Did that 18650 that was charged to 4.9 volts have any symptoms of being damaged?? Both batteries and LEDs often survive mistreatment, sometimes for quite a while.. And if an LED does not burn out immediately,, although it's life is hortened, it can still last for thousands of hours. Also, keep in mind that the life of an LED is considered ended when the output is reduced to half. That might be entirely adequate for some applications.

 

But how can I prevent the 18650 to charge over 4.2V, considering the one I removed was at 4.9V?

I would try connecting that TP4056 board as you described. I don't think pin6 is a controlling the charge but is the input that detects when the solar voltage is high enough to shut off the PIR circuit.

 

Did that 18650 that was charged to 4.9 volts have any symptoms of being damaged??

Yes, the capacity decreased substantially. When I tried to test it with a ZB2L3 battery tester, the voltage drop under load would end the test immediately. This doesn't happen with other healthy batteries I have.

 

I would try connecting that TP4056 board as you described. I don't think pin6 is a controlling the charge but is the input that detects when the solar voltage is high enough to shut off the PIR circuit.

I think I can confirm this. When the voltage at the pin falls down to about 1V (as I gradually cover the panel), the PIR becomes active and starts triggering the LEDs.

Maybe I can add the TP4056 like this?

 

Maybe I can add the TP4056 like this?

That's how I would connect it. Then I would change the default 1 amp charge rate to the lowest setting of 135ma if that's all the solar panel can deliver.

 

Thank you, I've changed the resistor to 10K according to the datasheet of the TP4056.