Re: Canon PowerShot A710 IS backlight

In article <7g7m93t27mj40etofl2ti6uts1fdchlff1@xxxxxxx>, Franc Zabkar wrote:

On 15 Jul 2007 23:07:24 GMT, u_1061771155@xxxxxxx wrote:
The driver circuit is somewhat hard to trace. (Anyone know of
a source for a schematic... ;-) So far I've traced it back
to a SOT23 sized 5-pin thing:

'large'
_ _____ _ inductor
drive?->--|_| |_|b-------@@@@@-----+------->
| XH |_ _|_ backlight
| 6 |_|GND 'tiny' ___ ?-->
_| G4 |_ cap |
a|_|_____|_|Vcc GND

I think "a" connects to "b" under the chip.

Could "a" connect to the junction of the cap and inductor, in which
case the chip may be a self contained, fixed voltage, PWM regulator
with integrated MOSFET, flywheel diode, oscillator, and error amp. "A"
may be the feedback pin for the error amp, ie this pin may sense the
regulator's output voltage. The "drive" may be an on/off pin.

Hi Franc,

"a" to "b" measures 0.01 ohm (with my Cap ESR/low ohms meter). By
comparison, across the inductor measures 0.23 ohm.
So I don't believe "a" goes to the other inductor/cap junction.

However I do suspect the chip contains what you describe. I found
numerous 5-pin SOT23 led driver datasheets, though all seemed to
have an external diode in a classic "boost" configuration and I
don't seem to have one.

Do you measure a diode drop between GND and "b" (anode to GND)? If so,
then this could be an internal flywheel diode.

It does measure like a diode, anode to GND, just as you describe.

I believe that the backlight could be a white LED type. This would
make the chip a down-converter which is consistent with the placement
of the inductor and an [internal] flywheel diode. Or maybe it's some
kind of switchmode constant current source ???

I haven't found any info on the A710, but FWIW the PowerShot A200/A100
both have LED backlights:
http://www.canon.com/camera-museum/tech/report/200207/report.html

- Franc Zabkar

Good pointer. Thanks.

You are indeed correct. I tried some DC (with 10K series resistor)
into the backlight but nothing happened. I then managed to peel
the backlight unit apart (two very interesting plastic films in
there with odd optical properties - diffraction gratings maybe)
and got the back layer out of the casing. Under a silver sticky
tape along the edge, I find two miniscule SMT white leds on the
flex. The LEDs work fine. However I've discovered a hairline break
in the flex cable wiring, near the connector, both the gold and
the underlying copper have been 'unplated' from it. A fiddly job
but I managed to bridge it with some fine wire.

I'll investigate the circuit a bit more before I reassemble,
as I don't understand the apparently open 2nd backlight terminal
on the PCB. It's a bit of a *** to assemble and disassemble,
since this means unsoldering wires, so I'd like to avoid
reassembling it and then finding it still doesn't work...

Thanks for your help.
Mike.

.

Quantcast