Samsung LN32B540P8D LCD panel or bad T-CON?

[budm]

I have this TV inline waiting for new panel for months, I looked at the pictures on the screen after warm up for about an hour, it will have lines, double images, slow motion similar to bad TAB bonding like SONY/SAMSUNG panel with bad bonding problem, did not have the time to take the panel apart to inspect until now.
So today I decide to take it apart to see if i can fix the TAB bonding problem that I have done on my 52" SONY (http://s807.photobucket.com/albums/y...20KDL-52VL150/).
The panel has two tabs, pressing on the them made no difference.
I then cool down the whole T-CON board, then the problem went away, after 5 minutes, the problem comes back again, then I try localize cooling, narrow it down to the the area around 3-diode arrays D22. D23, and the caps close by.
Then I spray the Q-tip with cold spray so I can really narrow down to the single component, touch D23, then the problem goes away.
Remove the D23 and I can see that the 3 Diodes are connected in series, same for D22. D22 is connected in series with D23, so basically 6 Diodes are connected in series, the Cathode is connected to ground, the Anode (top of the Diode string is connected to pin 5 of IC322 (i7822?).
I did not find the spec for the IC, but I think what happen is that the Vf of the Diode will go down when it is warm up, and goes up when it is cool down.
So I use 6 1N4148 in series and install them in the circuit, power up the TV, the pictures look normal, heat up the T-CON board with hair dryer, it still look normal.
TV has been running for about 2 hours so far with great pictures.
It is so hard to be 100% to be sure if it is T-CON or LCD panel is bad.

http://s807.photobucket.com/albums/y...20LN32B540P8D/

[retiredcaps]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

I did not find the spec for the IC,

I believe KAD is



Found via

[budm]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

Thanks, I found the Diode data but not the i7822 that the Diodes are connected to.

[retiredcaps]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

Thanks, I found the Diode data but not the i7822 that the Diodes are connected to.

If you mean the 8 pin IC (IC322), then I think it is an 8 pin mosfet.

http://pdf1.alldatasheet.com/datashe.../SI7820DN.html

[budm]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

The number as printed on the IC322 is i7822

[tom66]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

Interesting diagnosis! I would not expect those diodes to cause that problem. Perhaps they are part of the charge pump for generating the +30V TFT ON voltage?

[budm]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

If I bypass one of the 6 Diode in the string I replaced, the screen will show bad pictures again. I do not see why they use two Diode arrays and hook them all up in series, strange design. It looks like they trying to make something around 2.5V Vref, I would have used TL431. Need to find the spec for that i7822 IC.
TV is still running fine with 6 1N4148.

[tom66]

Re: Samsung LN32B540P8D LCD panel or bad T-CON?

Yes, well, when you're the engineer who figures out how to cut $0.10 from every LCD shipped, you get praise. Since it's a high load impedance (you can't really use such a reference with heavy load), and probably imprecise (diode Vf very variable with temperature, current and individual diode), they could possibly even use a resistor divider from the 3.3V rail, or a zener diode.