Update 7/12/18 - After a few day/night cycles the ni-cad is starting to stabilize. The first cycles had the battery discharging all night due to it being fully charged but this morning the led was out yet with just the first hour of sunlight, the battery charged enough to get the led lighting again when I covered the solar cell. So... tonight I will see how long the battery will discharge. I will be content if it will work till past my usual bed time of 9 or 10pm. I will be very happy if it will last past 2am when I usually wake up for work :-)
The bear is alive.....sort of.
This is the circuit being tested except that the solar cell voltage is about 2 volts and the diode used is a 1n914. The Joule Thief is connected where the led is indicated and everything seems to be working well.
Once I am sure everything is good after a few day/night cycles then I will reduce the circuit to a size that will fit inside the bear. The charging/switching circuit in the base and the Joule Thief in the lamp housing.
Did a bit of learning today by wiring up my first Joule Thief circuit. I never messed with these little circuits before as I never had a need to use one. This circuit does work first time and lights up the bears led using the 1.2 volt ni-cad battery.
The led is about as bright as when we first bought the bear a few years ago so that is a good start. Next will be the actual charging and switching circuit.
Here is the Joule Thief circuit used.
I used a FT50-43 core with 17 turns of number 28 enameled wire to start with.
First, the battery does take a good charge after connecting it to my variable power supply and a series resistor to charge at about 25ma overnight C/20 more or less.
Second, When I place the battery back into the circuit and switch on there is constant voltage on the led wires but no light, very strange. The led is not lighting at all and even putting full light onto the solar cell, the circuit does not cut off the voltage to the led as it should.
Assumption number one ( yeah, I know not to use that word )....
Led is bad for what ever reason...
Number two, If the led worked then It would be constantly drawing current and the battery would never charge...
Not sure what type of white led is being used here but the output voltage is only at the 1.2 - 1.4 volt level
The circuit board only has a resistor and single special 4 leg semiconductor package and the print is so small that I can not read it and no magnifier around :-)
Time to think about rolling my own transistor circuit and perhaps a Joule thief to boost the voltage up so I can use a different led...
I should say that the battery did it.... I think :-)
I measured the charging voltage in full sun at about 2 volts. The battery was down to 0.1 volt from being in storage for a year or so. After charging the battery all day in strong sun I have measured the battery voltage at 1 volt and it should be 1.2 volts.
I will have to go and get a replacement Ni-Cad and see if that was the only problem. All other connections seem to look good.