cobbling together a ni-cad charger

[kell]

I have a cordless floor sweeper that's missing the wall wart power
supply it came with. Inside the sweeper there is a 7.2 volt nicad pack
with just a power resistor to limit current to the battery when
charging (very primitive). The resistor is something like an ohm, and
the size is a watt or two. I don't know what voltage the wall wart
was, but it was probably unregulated. The sticker on the sweeper says
that when new, and charging the (empty) battery for the first time, to
let it charge for 16 hours.
When experimenting I hook the sweeper up to wall warts that have 15 or
20 volts open circuit things get hot -- the resistor in the sweeper,
and the wall wart.
I have a 7808 8-volt regulator I could wire up to a 12 or 15 volt wall
wart. My question, is 8 volts enough for charging a 7.2 volt nicad
pack? A really slow charge is good, this thing will sit in the hall of
the building where I live and probably just stay on the charger all the
time. If 8 volts is too low, I can get whatever voltage I need by
putting a diode or diodes in the ground terminal of the regulator to
bump the voltage up, or use a voltage divider like on a 317... I'm
really asking what fixed voltage would be best, with a 1-ohm resistor
in series, to leave a 7.2 volt nicad pack on all the time and not
overheat; it's okay if it takes a whole day to charge.

 

[spudnuty]

I have a circuit that uses a light bulb as above and a LM317 voltage
regulator to limit charging current to 100ma and adjust the cutoff
voltage as required.
Richard

 

[kell]

It's a little unclear whether spudnuty used the 317 in the
constant-current config or in the usual voltage-regulating config and a
bulb in series with the output.
Mikes suggestion is the simplest but I want a setup where the charging
current drops off as the nicads voltage rises. That's why I was
thinking about using a comparatively low 8 or 9 volts, so that as the
batteries rising voltage approaches that, the voltage across the
current-limiting element (bulb or power resistor) will fall very low,
causing the charging current to diminish. Actually I know this is not
the best way to charge nicads; it is more how you would charge
lead-acid batteries. Maybe I should tape a thermistor to the battery
and use that to turn off the charging.

 

[kell]

that timer sounds like a good idea.