"Dick Smith" 55" LED backlight issue - model GE6880

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

I just had the multimeter on current setting in series with the batteries and led being tested.

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Tried different angles etc and it seems that it definitely turns off a frame or two earlier than the rest.

If that is the cause I still don't know why the led open/short fault hasn't triggered though. It seems extremely odd to me that the chip would keep driving the transistors etc even after such a "fault"?? Also if this were the cause why would it drive the VFB pin low? Isn't that in essence trying to pull even more power out of the primary side controller (and hence triggering that controllers power-limiting shutdown)?

[SLK001]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

The fault doesn't occur until the LED switches off (becomes open circuit).

I've had new LEDs that seem to oscillate - would be on for a 1/4 second, then off for a second. It was a new diode and I just replaced it with another.

The LEDs are fairly cheap... What do you have to lose by replacing it?

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

The fault doesn't occur until the LED switches off (becomes open circuit).

I've had new LEDs that seem to oscillate - would be on for a 1/4 second, then off for a second. It was a new diode and I just replaced it with another.

The LEDs are fairly cheap... What do you have to lose by replacing it?

I get what you're saying, and I will try swapping in a new LED. Its just not triggering the open led fault of the controller that's bugging me and making me think the fault is elsewhere. Even leaving the LEDs unplugged doesn't even trigger the open led fault:/

For testing purposes does the replacement LED have to be identical? I've got a spare (definitely working) strip out of a different TV that I can pull some LEDs off to swap over, would they do to indicate if it is that led?

[SLK001]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

For testing purposes does the replacement LED have to be identical?

That usually works. Just be aware that some LEDs have large pad on the CATHODE and some have the large pad on the ANODE. Although one can be flown in on the other's pad, it really isn't a long term solution, as the heat isn't removed as well.

For testing, I have put down a leaded white LED to test out a TV.

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Before i go popping off lenses and desoldering LEDs, is there any way to use the test points (they were nice enough to put one on either side of each led)? Would soldering in a second LED with wires to the test points work for testing purposes? Or perhaps a resistor, just something to test if this LED is the cause before doing something potentially damaging .

[SLK001]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Apply 3 volts to the suspect LED and measure the current. Hold the voltage for 10 to 20 seconds. Since this LED is the one shutting off prior to the rest of the panel, it should shut off by itself.

Use a hair dryer to warm up the glue holding the lens and it should pop off fairly easily. Use BudM's method of removing the LED if you don't have two soldering irons. If you don't over do it, there is little possibility of doing damage.

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Well it didn't turn off for over a minute..
But then I went to test the next one in the chain (to check the current was still close) and got the wrong test point so it was putting the 3v over 2 LEDs, and oddly the 2nd (good) one illuminated very dimly..
Turns out the 1st led has indeed failed, has a measurable resistance across it, ~300kOhm. Probably failing totally at peak current output.

Still no idea why it isn't triggering the drivers protection though.


Will try swapping in a good LED tomorrow, hopefully this led is all it is.

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Use a hair dryer to warm up the glue holding the lens and it should pop off fairly easily. Use BudM's method of removing the LED if you don't have two soldering irons. If you don't over do it, there is little possibility of doing damage.

Is there any special glue or anything recommended for re-attaching the lenses?

Also what is BudMs method (or a link to where it is)? Tried to find it but couldnt see any thing obvious.

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Swapped out that LED with a good one off another string (actually tried 4 off two different good strings- just to be sure I hadn't damaged them in removal/replacement). ANNDD.... its still exactly the same as before - everything is the same as far as I can tell, measured the various pins and as far as DMM measurements go its all operating exactly the same as with the "bad" LED.

I have also tested all the other LEDs for resistance, none of the others have a reading on the 2MOhm scale. Also none of the others turn off quickly or before the others etc on my video.



This one is really stumping me. I would really like to isolate which part is bad - the LEDs or the PS board.


I was thinking as a way to do this:

Upping the value of the ISET resistors (circled in picture), to kick in the current limiting part of the driver at a lower current.

I know every LED works for at least a few seconds at 100mA (battery test showed that) so if I up the resistance to 20 - 30 Ohm on each driver branch (its currently at 8.5 Ohm total combining the two resistors - this differs from the schematics 5Ohm total with two resistors, I suspect because two driver circuits have been paralleled into each string as opposed to the one per string in the schematic) per branch would drop the current limit sufficiently to get the LEDs to work for at least a few seconds (if the power board is good).


If the LEDs do work then its a bad LED issue. if not its a power board issue.




Is that a good thing to do, or have I missed something and doing that would be terrible?

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

To revive this thread a bit with new information:

After putting this into the "too hard" category for awhile I was lucky enough to come across an identical TV with an "intermittent fault" that turned it off after a few hours and needed a few minutes unplugged before turning back on, decided it was worth the risk to buy it cheap as a diagnosis tool and potentially fix 2 tvs at the same time so got it.


so now for the updates:


Took the power board from that TV and plugged it into the backlight assembly from the original TV and and it works!

Then to confirm I put the PS board from the orignial TV into the "new" TV and exact same issue, just a single blink.

So from this i believe it is relatively safe to conclude that (both) the backlight assemblies are working correctly and that the fault is now confirmed to be in the power board.



I should also note that the second TV appears to have a pre-failure version of the same fault. After turning off (It took almost 2 days straight use in a hot room before this happened, then it needed to sit for a few minutes before coming back on again) if I shine a torch on the screen I can see that the screen is still running but no backlight, and if I turn it back on immediately it just gives the exact same brief flash. of course if this one sits for a few minutes it all works again.


Does this give anyone any more insight into what the failed component may be? keep in mind that I've already swapped the OB3364 controller over so its unlikely to be that!

I was considering just ordering a bunch of parts and swapping almost every component in the LED primary and secondary sides over - would still be cheaper than buying a replacement board (which would then have to be modified anyway seeing as this one had been modified to run the 55")!

tl;dr LED backlights confirmed as good, fault is in the PSU - still need help isolating the fault.

[mmartell]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Use hot air, a wand if possible so you can isolate individual components.

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

SOLVED!

the culprit was DIODE "D112" on the schematic.

Despite showing the correct forward voltage drop for the diode (.46V) under a diode test, and no measurement in reverse, it has ~1.3kOhm resistance across it. would have found it far earlier had I thought to do a resistance test over the diode despite the seemingly good diode test result!



now, seeing as the 2nd PSU is failing in the same way this diode may well be a weakness in the 55" implementation of this circuit.

So I was thinking that perhaps it is underrated.

would it be prudent to choose an uprated diode?

If so is it the average current or the blocking voltage (or both) that should be uprated? my local electronics store sells FR607 (6 amp version of the part) would getting that be a good idea to hopefully overcome this possible weakness?

[budm]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

So the snubber diode D112 is bad, good find, may be it is the brand they are using. What is the Voltage rating?

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

its a FR307 - 1000V maximum DC blocking, 700V max RMS voltage. 3A rating.

I don't recall having seen the manufacturers logo before, its a black "R" (with white border) with a smaller white "S" or maybe "5" inside it (creating the holes at the top and bottom of the R) - don't know if that description makes any sense!

perhaps its just a cheap knockoff and doesn't actually meet the specs of the official FR307.

[saqi4u]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Hi am new here i ahve same model haire 55 with tv5001-zc02-01 led backlight not working, its just flash when power on then gose blank.
on Both LED con anode has 110V if i test with ground to anode'
if i test on both LED con from anode to cathode 0 voltages
it mean -v have problem plz help me

https://www.badcaps.net/forum/attach...4&d=1461991048

sorry my English is very poor

[rockit711]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

check D112 (large diode on the primary side of the LED transformer circuit) it was a "FR307" on mine. it resulted in the same "flash" then nothing as you describe. May need to be de-soldered to get a good test on it. Also do a resistance check on it in case it has partially failed (as mine had).


If that isn't it then I'd recommend looking at pin 9 of the OB3354 chip (its the 16 pin SMD one just to the left of the LED connectors you have marked in your picture).

Pin 9 is the fault pin, it will trigger if the driver detects a failed LED (short or open) and shut down the backlight.

if pin 9 has triggered (high output from memory) then it means the fault is almost definitely a failed LED(s) in the backlight and you will have to isolate the failed LED and replace.

[streetford@hotmail.com]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Hey Rockit711,
I am fixing the same tv for a friend. I also need to replace D112 as mine has the FR307 part in it. Did the FR607 do the job ok or did you go with a different one?? Thanks... if not for your determination no one would be fixing these

[joezzz]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Hi guys, I have the same scenario .... I went ahead and changed over the FR307 to no avail. Back light goes out after about 20 minutes ,,, then turned on again it goes off after approx 12 minutes,,,,then again after 1 minute .....then just flashes all the time voice works ok ..



i have used a fan for cooling and have since soldered all the connections on the board to no avail . I bought this set on the 29th of july 2016 now this...any advice guys..

PS strangely it was turned on by my niece and worked for hours until I turned it off..


any ideas guys

[FamilyMan]

Re: "Dick Smith" 55" LED backlight issue - model GE6880

Well, for what its worth, this is my experience.
I collected one of these off the kerbside cleanup, no backlight. Went through and tested each backlight LED in turn and found two bad ones. Since it is fed essentially by a current source, I just shorted out those two with a piece of wire and put it all back together and it worked for 12 months. I put up with the fact that there are two widely separated slightly dim areas of backlight. I'm the only one who can see it. It worked that way happily for 12 months or so.
Then the backlight went out again and I immediately suspected another bad LED, so stripped it down and it all checked out ok. I even went to the trouble of creating a new test 140V power supply, and all the LEDs came up just fine.
Anyway, to cut a long story short, went through all of the issues in this forum to no avail. In my case it turned out to be U204, on the schematic as a TL432 2.5V shunt regulator. But mine had fitted a MB432A https://pdf1.alldatasheet.co.kr/data...TC/B432AM.html, which you might assume is a TL432 equivalent. But no, it is a 1.25V part, not a 2.5V. I twigged to this because the reference pin was sitting at 1.3V and the part was conducting, so I assumed it was stuffed, but it all checked out ok. Anyway, this causes no end of havoc with the feedback because the voltages are out by a factor of 2. The wonder is that it worked at all, but I suspect the main feedback is via the OB3354 mechanism which isn't so voltage reference sensitive. However, in the initial case I think the backlight voltage was low by a factor of two and so I think it was more prone to triggering the fault and killing the backlight.
So, without a proper TL432, I retained the MB432A and fitted 100k to R237 (vacant), put 2k2 in parallel with R235. and replaced R232 with 47k to make the feedback values correct for a 1.25V reference. I have no idea if that was all necessary, but that is what I did. And with all of the boards back to make the power on and adjust signals correct, up it came.
YMMV.