Hello,
I would like to replace a faulty backlight LED in a TV. The original backlight strips seem a bit difficult to get, and I want to try if I can get away with just replacing the LED.
LED strip type: Sharp A668WJ-9X
The question I have is how to identify the LED used on that strip. It is a 3030 type (though closer to 2.9mm according to my caliper). I could see two LED die under the microscope in the chip. What seems unusual is the electric characteristics. I measured the following using a DC power supply:
5.6V - 5mA
6.0V - 11mA
6.5V - 22mA
7.0V - 32mA
When I go look on digikey, all 3030 LED with 6V to 6.4V nominal voltage seem to be rated 100mA - 200mA at that voltage. Any insight on typical and a-typical LED used in those strips would be appreciated.
Also, the power supply for those strips is quite interesting. I don't know much about SMPS topologies, but this seems to be a sort of 4x stacked flyback type, where each "slice" of four individual transformers (no magnetic coupling), primary sides serially "stacked", generates two channels of LED strips, i.e. 8 rows of LED driven by individual lines. The four serially stacked transformers are driven by the secondary winding of a first stage towards the "hot" side. I will attach a quick diagram.
Besides the fact I find this power architecture interesting and would like to know how this is properly called, I am also considering the effects of introducing imbalance if I replace one LED by a type that doesn't match well. I have already replaced the faulty one by a 235 Ohm resistor, and it seems it is driving a current, i.e. I got the same voltage and brightness on other LEDs, but higher voltage only across the resistor. I have to check again the complementary strip, i.e. the one driven by the same transformer, but I believe it is driven as all the other strips. Measured with a digital multimeter, it shows 7.3V across one LED at high brightness when the TV is turned on.
Thanks for your input!
Dodano po 8 [godziny] 10 [minuty]:
I made a funny experiment. I took some 3535 LED from an old bicycle light, and soldered two in a row on the backlight strip. I knew from testing that at lower voltages, the current is lower compared to the original. So I expected it to drive quite high voltages at the replacement. It turns out that at highest brightness (a few seconds at boot time), the voltage across the replacement is actually 300mV lower than on the originals. On lower brightness, the voltage across the replacement is 200mV higher than on the originals. In other words, the replacement seems to have a flatter characteristic curve. I am tempted to put the TV back together to see if and how the replacement affects picture quality. The white is certainly a bit warmer than the originals - then again, there is about 100 of them in this 32" TV, quite a dense population. The real question however is how long will the replacement last, and what happens if it fails, with a failure to short for example.
I would like to replace a faulty backlight LED in a TV. The original backlight strips seem a bit difficult to get, and I want to try if I can get away with just replacing the LED.
LED strip type: Sharp A668WJ-9X
The question I have is how to identify the LED used on that strip. It is a 3030 type (though closer to 2.9mm according to my caliper). I could see two LED die under the microscope in the chip. What seems unusual is the electric characteristics. I measured the following using a DC power supply:
5.6V - 5mA
6.0V - 11mA
6.5V - 22mA
7.0V - 32mA
When I go look on digikey, all 3030 LED with 6V to 6.4V nominal voltage seem to be rated 100mA - 200mA at that voltage. Any insight on typical and a-typical LED used in those strips would be appreciated.
Also, the power supply for those strips is quite interesting. I don't know much about SMPS topologies, but this seems to be a sort of 4x stacked flyback type, where each "slice" of four individual transformers (no magnetic coupling), primary sides serially "stacked", generates two channels of LED strips, i.e. 8 rows of LED driven by individual lines. The four serially stacked transformers are driven by the secondary winding of a first stage towards the "hot" side. I will attach a quick diagram.
Besides the fact I find this power architecture interesting and would like to know how this is properly called, I am also considering the effects of introducing imbalance if I replace one LED by a type that doesn't match well. I have already replaced the faulty one by a 235 Ohm resistor, and it seems it is driving a current, i.e. I got the same voltage and brightness on other LEDs, but higher voltage only across the resistor. I have to check again the complementary strip, i.e. the one driven by the same transformer, but I believe it is driven as all the other strips. Measured with a digital multimeter, it shows 7.3V across one LED at high brightness when the TV is turned on.
Thanks for your input!
Dodano po 8 [godziny] 10 [minuty]:
I made a funny experiment. I took some 3535 LED from an old bicycle light, and soldered two in a row on the backlight strip. I knew from testing that at lower voltages, the current is lower compared to the original. So I expected it to drive quite high voltages at the replacement. It turns out that at highest brightness (a few seconds at boot time), the voltage across the replacement is actually 300mV lower than on the originals. On lower brightness, the voltage across the replacement is 200mV higher than on the originals. In other words, the replacement seems to have a flatter characteristic curve. I am tempted to put the TV back together to see if and how the replacement affects picture quality. The white is certainly a bit warmer than the originals - then again, there is about 100 of them in this 32" TV, quite a dense population. The real question however is how long will the replacement last, and what happens if it fails, with a failure to short for example.
