LED Backlight current question
*long story! Skip to second part for actual question*
Good day folks. I was messing around with a couple of TVs that got left behind in the shop, essentially being knackered in one way or another and I found two which are almost identical and each have a different issue: one has a dead power supply and the other a dead main board, so I thought I'd combine the two to make at least one of them functional, which got me thinking about one thing which I'll discuss later on. The main boards aren't quite direct replacements, not matching physically for one and secondly the panels are different sizes: one is a 24, the other a 22 incher. The PSUs however are virtually identical outside maybe their brand and the position of the components, so it was easier to adapt the PSU, instead of the main board and messing with firmwares and trying to matche the panels and just....naaah.
The only difference is that one panel had only two wires for the backlight (+ and -) and the other one had 6 (2 +, 2- and two dummies which we don't care about). Unfortunately the functional PSU belonged to the 2 wire panel. I pushed on with this adaptation because I noticed despite it having more wires, they were nothing more than simply paralleled, so all - were bunched together and all + were bunched together, so I maintained that arrangement and soldered the red bunch to LED+ and the black bunch to LED-...while the TV fired up with the replacement PSU, the backlight didn't - not even a spark or something...it just sat there. I then tried multiple combinations of those wires, removing one set and leaving just the other alternatively, but still nothing. I managed to get it going however: guess what the big secret was....the polarity was reversed, in other words RED was negative and BLACK was positive all along....I would've never figured it out had I not pulled the datasheet for the panel which thankfully is an LG and has a comprehensive one. So I soldered black to + and red to - as counterintuitive as it was, but only had one wire pair connected (the other was floating - we're talking about the 22 incher which has 2 pairs of wires). This time the backlight did flash on for a second upon applying power and 3v to the BL-ON pin and once I connected the second pair as well, the backlight remained on properly.
This got me thinking however, despite it not being fully related to what I did there today: assuming the LED driver is a constant current source designed for a specific type of panel with a specific current draw, assuming I try adapting a CC source for a smaller panel with a smaller current draw, what would I need to change ? I initially thought I'd HAVE to perform such a hack to enable the smaller panel to be run by the driver of the bigger panel (having less LEDs and thus smaller draw), but it turned out it wasn't necessary after all. Assuming it was, what would it come down to ? The current sense resistors ? When talking about CS resistors on general, would an increase in the value result in a higher load, tricking the PSU into running the smaller load, or is a DECREASE in value necessary for this purpose ? I messed around with Ohm's law and I THINK an increase in resistance results in the PSU detecting a higher voltage drop across the resistor, hence a higher load. If the PSU is expecting a value of X volts dropped on that current sense resistor, a LOWER value than expected would be like a lower or missing load, like in the case of a disconnected or open LED strip (which is why my PSU shut down back there when I only had one string of LEDs connected - when I connected the other one, the PSU stayed on). Assuming I HAD no second strip and wanted to run just the first one, how would I achieve this ?
Think of it like this: a 1A load across a 1ohm sense resistor would result in a 1v drop, whereas if I had the same 1A load but through a 2ohm resistor, this would result in a 2v drop....V=IxR ? Assuming it's 2v the comparator expects, it stays on. With the smaller value, the drop is lower and the comparator detects an "open lamp" fault....is this close to being a correct assumption ? Cheers and thanks.
Good day folks. I was messing around with a couple of TVs that got left behind in the shop, essentially being knackered in one way or another and I found two which are almost identical and each have a different issue: one has a dead power supply and the other a dead main board, so I thought I'd combine the two to make at least one of them functional, which got me thinking about one thing which I'll discuss later on. The main boards aren't quite direct replacements, not matching physically for one and secondly the panels are different sizes: one is a 24, the other a 22 incher. The PSUs however are virtually identical outside maybe their brand and the position of the components, so it was easier to adapt the PSU, instead of the main board and messing with firmwares and trying to matche the panels and just....naaah.
The only difference is that one panel had only two wires for the backlight (+ and -) and the other one had 6 (2 +, 2- and two dummies which we don't care about). Unfortunately the functional PSU belonged to the 2 wire panel. I pushed on with this adaptation because I noticed despite it having more wires, they were nothing more than simply paralleled, so all - were bunched together and all + were bunched together, so I maintained that arrangement and soldered the red bunch to LED+ and the black bunch to LED-...while the TV fired up with the replacement PSU, the backlight didn't - not even a spark or something...it just sat there. I then tried multiple combinations of those wires, removing one set and leaving just the other alternatively, but still nothing. I managed to get it going however: guess what the big secret was....the polarity was reversed, in other words RED was negative and BLACK was positive all along....I would've never figured it out had I not pulled the datasheet for the panel which thankfully is an LG and has a comprehensive one. So I soldered black to + and red to - as counterintuitive as it was, but only had one wire pair connected (the other was floating - we're talking about the 22 incher which has 2 pairs of wires). This time the backlight did flash on for a second upon applying power and 3v to the BL-ON pin and once I connected the second pair as well, the backlight remained on properly.
This got me thinking however, despite it not being fully related to what I did there today: assuming the LED driver is a constant current source designed for a specific type of panel with a specific current draw, assuming I try adapting a CC source for a smaller panel with a smaller current draw, what would I need to change ? I initially thought I'd HAVE to perform such a hack to enable the smaller panel to be run by the driver of the bigger panel (having less LEDs and thus smaller draw), but it turned out it wasn't necessary after all. Assuming it was, what would it come down to ? The current sense resistors ? When talking about CS resistors on general, would an increase in the value result in a higher load, tricking the PSU into running the smaller load, or is a DECREASE in value necessary for this purpose ? I messed around with Ohm's law and I THINK an increase in resistance results in the PSU detecting a higher voltage drop across the resistor, hence a higher load. If the PSU is expecting a value of X volts dropped on that current sense resistor, a LOWER value than expected would be like a lower or missing load, like in the case of a disconnected or open LED strip (which is why my PSU shut down back there when I only had one string of LEDs connected - when I connected the other one, the PSU stayed on). Assuming I HAD no second strip and wanted to run just the first one, how would I achieve this ?
Think of it like this: a 1A load across a 1ohm sense resistor would result in a 1v drop, whereas if I had the same 1A load but through a 2ohm resistor, this would result in a 2v drop....V=IxR ? Assuming it's 2v the comparator expects, it stays on. With the smaller value, the drop is lower and the comparator detects an "open lamp" fault....is this close to being a correct assumption ? Cheers and thanks.
Comment