In 2007, my brother got an LG 26LC7R HD-ready TV, and we shelved the still-working(!) 20 year-old CRT set. The new TV had analog and two HDMI inputs. Now, ten years on…

Problem 1

The LCD TV started exhibiting delayed video functionality after power on, which progressed from two minutes to 45 minutes over a few months; audio remained normal. At power on, a white screen (without OSD) was gradually replaced by vertical bands of finer lines (Figure 1), followed by normal video after the delay. This was common for all video inputs. It looked like some faulty component warmed up during the delay and recovered.


Figure 1 The vertical-lines problem


Figure 2 Layout of the four main boards

The company technician diagnosed it as main board problem, with no spare boards available. Well, no EE with DIY leanings can stand a dysfunctional TV with bright CCFLs and otherwise good build for long. Armed with a multimeter and some forum suggestions, I opened it.

The SMPS board had three bulging 680µF/35V capacitors near the heatsink (Figures 3 & 4) on the 24V 3A CCFL inverter supply. Heat from the heatsink doubtless helped vaporize the electrolyte. All large electrolytic caps were replaced, and output voltages measured on the SMPS connector were normal, but the first problem continued.


Figure 3 Bulging capacitors


Figure 4 Closer view of the caps

Problem 2

After working for 45 minutes in HDMI mode, the video became a still picture, fading into white, and it seemed to run hot. Again, audio was okay. Switching off for a few minutes to cool restored operation immediately. With power unplugged, I felt hotspots on the secondary SMPS heatsink, main board ICs, main processor (hottest), and HDMI switch IC.

To confirm thermal shutdown as a possible cause, I set up a fan to cool the open TV’s main board. It operated continuously. The main board had a number of LDOs and (aging) SMD caps, though all voltages were okay. Could unstable regulator outputs be the cause of the heating? I recalled EDN articles cautioning about LDO instability, and added electrolytics at the outputs as shown in Figure 5 to the existing SMD ones. However, they did not change anything, and were later removed.


Figure 5 Caps added and removed

[Continue reading on EDN US: Measuring & solving it ...]


Related articles:

is an EE with experience in embedded systems applications and EMC.