I just got a Samsung PS43E450 from a friend for doing a favour for him.
It was dead, with no picture. Vs and Va present, but shut down shortly after and control board goes into protect mode (5-2 blink)
Scoping the Y-main waveform revealed the set was shutting down after the first sustain field. The Y-main waveform was distorted badly, there was no Vs portion and only the negative part was present. The X-main waveform, for the few cycles the set was active for, appeared OK.
This set uses a very cheap and very modern design where the energy recovery path is on the Y-main and the X-main only handles the Vs and GND transistors. Even the Ve bias path is on the Y-main, which is unusual.
There were no shorted transistors. So I began checking voltages. Vscan and Vsch were both present, as was Ve. And Vs was OK for the few seconds the set stayed on. As was Va. Everything appeared OK voltage wise until the shutdown.
I was about to remove transistors to check individually, in case one had failed somehow. However I decided to look at the board more closely first.
I found that the 15V_F supply, which sits on top of the Y-main output (VFG in Panasonic terms, YOUT in Samsung terms) had a 5 ohm short to ground. I was able to locate this supply as it is visible as a tap coming off the main switching transformer on the board. (This transformer is responsible for 15V_F, another 15V rail, Vscan, Vsch and Ve. I would not want to be the engineer who designed that, balancing those loads must have been a bitch!) Without the 15V_F supply, no high-voltage output can be properly generated. This includes the Y-ramp (Yrr, Yfr) and Ys parts of the waveform. It will also cause a shutdown because the ERC path voltage will drop below a threshold (same cause as Panasonic SOS6, though the mechanism on a Samsung set is not well understood for me).
The board makes it easy to cut jumpers to isolate different sections - a nice design, although I suspect this is not deliberate but a function of the 1-layer layout. In any case, by cutting jumpers, I was able to locate the short to the section of the board that drives the YS switch for bringing the Y-main output to Vs. I thought that the driver might be bad, but each driver is fed via a 15 ohm resistor. So a 5 ohm short would not be possible, it would have to be at least 15 ohms (I checked the resistor to make sure, and indeed there was 5 ohms to YOUT on one end, and 20 ohms to VFG on the other end, a sure indicator that the short is on the 5 ohm end.) I checked an SMD ceramic and it was OK. The only thing left was a small ceramic through hole cap, value 1uF (code 105) and position C5030. Sure enough when I removed it the short went. And the cap tested 5 ohm shorted.
After removing the cap and replacing it with a SMD ceramic equivalent, the TV works now. Although there is a little maldischarge on some pixels in shaded areas, though it is quite minor. I will check the Vscan voltage later in hopes of fixing this as well.
It was dead, with no picture. Vs and Va present, but shut down shortly after and control board goes into protect mode (5-2 blink)
Scoping the Y-main waveform revealed the set was shutting down after the first sustain field. The Y-main waveform was distorted badly, there was no Vs portion and only the negative part was present. The X-main waveform, for the few cycles the set was active for, appeared OK.
This set uses a very cheap and very modern design where the energy recovery path is on the Y-main and the X-main only handles the Vs and GND transistors. Even the Ve bias path is on the Y-main, which is unusual.
There were no shorted transistors. So I began checking voltages. Vscan and Vsch were both present, as was Ve. And Vs was OK for the few seconds the set stayed on. As was Va. Everything appeared OK voltage wise until the shutdown.
I was about to remove transistors to check individually, in case one had failed somehow. However I decided to look at the board more closely first.
I found that the 15V_F supply, which sits on top of the Y-main output (VFG in Panasonic terms, YOUT in Samsung terms) had a 5 ohm short to ground. I was able to locate this supply as it is visible as a tap coming off the main switching transformer on the board. (This transformer is responsible for 15V_F, another 15V rail, Vscan, Vsch and Ve. I would not want to be the engineer who designed that, balancing those loads must have been a bitch!) Without the 15V_F supply, no high-voltage output can be properly generated. This includes the Y-ramp (Yrr, Yfr) and Ys parts of the waveform. It will also cause a shutdown because the ERC path voltage will drop below a threshold (same cause as Panasonic SOS6, though the mechanism on a Samsung set is not well understood for me).
The board makes it easy to cut jumpers to isolate different sections - a nice design, although I suspect this is not deliberate but a function of the 1-layer layout. In any case, by cutting jumpers, I was able to locate the short to the section of the board that drives the YS switch for bringing the Y-main output to Vs. I thought that the driver might be bad, but each driver is fed via a 15 ohm resistor. So a 5 ohm short would not be possible, it would have to be at least 15 ohms (I checked the resistor to make sure, and indeed there was 5 ohms to YOUT on one end, and 20 ohms to VFG on the other end, a sure indicator that the short is on the 5 ohm end.) I checked an SMD ceramic and it was OK. The only thing left was a small ceramic through hole cap, value 1uF (code 105) and position C5030. Sure enough when I removed it the short went. And the cap tested 5 ohm shorted.
After removing the cap and replacing it with a SMD ceramic equivalent, the TV works now. Although there is a little maldischarge on some pixels in shaded areas, though it is quite minor. I will check the Vscan voltage later in hopes of fixing this as well.
Attached Files
Comment